clock menu more-arrow no yes mobile

Filed under:

This voting reform solves 2 of America’s biggest political problems

“Proportional” voting would reduce party polarization and the number of wasted votes.

House districts in Ohio, 2011-2013: Map.
The House districts in many states, including Ohio, are non-competitive.
Wikimedia Commons

This piece was first published April. In June, Rep. Don Beyer (D-VA) introduced the Fair Representation Act, which would implement the reforms Lee Drutman proposes here.

Back in October, Alec MacGillis penned a provocative New York Times opinion piece titled “Go Midwest, Young Hipster.” The argument was straightforward. Young educated progressives were fleeing their hometowns in places like Iowa and Ohio for thriving cities like San Francisco and New York City — but those moves had profound political consequences. New York and San Francisco are already deep-blue cities in deep-blue states, and so while the well-educated, left-leaning emigres might increase the margin by which a Democratic senator or Congress member might win, they would be unlikely to change the results.

Back in swing-state Iowa or Ohio, however, their votes might actually matter. If enough of them stuck around, they help the Democrats take the House and Senate.

It’s safe to say there will be no hipster invasion of the Midwest. But the question is not going away: Why do we accept an electoral system in which your vote is far more likely to shape Congress if you live in Des Moines than if you live in San Francisco?

The current system is unfair not only because it leaves many citizens on the sidelines in solidly Republican as well as solidly Democratic districts and states, but also because it undermines political accountability and turbocharges polarization.

Polarization is often described in terms of red states and blue states, but it is a significant problem at the Congressional-district level across all the states. It’s also a more complex story than is usually suggested: Gerrymandering, or the partisan redrawing of district lines — a frequent object of complaint on the left —- has undoubtedly helped make some districts more unshakably Republican. (Democrats play the gerrymandering game, too, but they have had less opportunity.)

But gerrymandering probably accounts for less polarization than is often suggested, relative to other important trends, most notably the disappearance of socially liberal Republicans and socially conservative Democrats. They once contributed to many more closely contested and therefore moderation-encouraging congressional elections.

Whatever the causes of polarization, there is a relatively straightforward solution to our current predicament that has been embraced by most advanced industrial democracies: proportional representation. There are many versions of this approach, but they all involve some way of electing multiple people, at once, to represent a region. In a proportional system, parties representing as little as 1 percent of the electorate can gain representation, though the most stable systems usually have a threshold percentage level to prevent truly marginal parties from gaining seats. The regions can be as large as an entire nation — but even when they are smaller they tend to be larger than the 435 tiny US congressional districts, each of which is run according to the “winner take all” principle.

Under a proportional system, if you want to live in a big, liberal city in a liberal state, you don’t give up the chance to make a difference with your vote. There is also very little possibility for consequential gerrymandering in proportional representation systems, since districts tend to be so big that there’s not much to gain from alternative line-drawings.

Perhaps most significantly, proportional representation makes third parties more viable. In the US system, many voters might prefer a third party, in theory, but in a winner-take-all scenario a vote for a third party is a wasted vote, since only the two major parties stand a chance of winning. As a result, most proportional systems have at least three major parties, often more. This produces a wider diversity of perspectives in the representative body, and more potential for bargaining across different issues.

Because more parties are competing for voters; because voters are more likely to feel like their voters matter; and because voters are more likely to have the chance to vote for a candidate they are excited about, proportional representation systems tend to have higher voter turnout.

We’ve gotten used to our winner-take-all approach to elections, but proportional representation needn’t be a pie-in-the-sky idea. A group called FairVote has proposed the Fair Representation Act, which would transform the patchwork of state-level congressional districts into a larger ones — typically with three to five members for each district. Members would be elected through a ranked-voting system —an additional reform that lets voters express their true preference while expressing a secondary preference for someone from among the more viable candidates.

FairVote’s proposal is constitutional — the Constitution offers states quite a bit of leeway in selecting representatives — but it would require national legislation to reverse existing law mandating single-member districts. The proposal has historical precedent, however: It would move us back to the multi-member districts that were once more common. There would still be 435 members of Congress.

How our current voting system fuels bitter political division

To understand why this solution makes sense, let’s look at the existing problem in more depth, exploring why the current single-member winner-take-all system is a key driver of our current polarization.

In the United States, the vast majority of congressional seats are solidly safe for one party or the other. It’s still early, but as of April 21, the Cook Political Report lists only 5 out of 435 congressional races as “tossups,” and puts only an additional 19 races in its next closest category — “leaning” toward one party or the other. In other words, only about one in 20 Americans lives in a place that appears likely to have a competitive House election.

This is not new. Since the 1980s, consistently about three in four incumbents have won with at least 60 percent of the vote. In other words, these races are not even close.

chart showing percentage of house districts where the victor won by 60 % or more. Brookings Institution, Vital Statistics on Congress

Political scientists first began noticing the decline in the competitiveness of congressional elections in the 1970s, At first, they attributed the trend to the ability of incumbents to deter challenges through constituency service, earmarks, campaign fundraising, and other forms of self-promotion. They solidified support in their own party and stole some voters from the other party.

But a new degree of party loyalty also began contributing to the shift. Starting in the 1990s, the share of party identifiers voting for their party’s candidate — a number that had been declining for decades — began to increase. Today, party-line voting for congressional candidates is back around 85 percent.

Percent of partisans voting for own party’s candidate in US House elections Brookings Institution, Vital Statistics on Congress

This happened for a few related reasons. For one thing, parties have become more internally homogenous. Starting in the 1970s, as ethnocultural issues — social and identity-related issues — became more central to our politics, socially conservative Southern Democrats and socially liberally Northern Republicans began to switch parties. This reduced the ideological overlap between the two parties in Congress. In turn, lack of overlap gave party leaders even more power to draw sharper distinctions between the two parties by shaping agendas and controlling messaging.

As differences between the two parties hardened, voters became increasingly likely to pick a side and stay with that side, contributing to safer districts and more polarization. Today, as noted, about three in four incumbents win with more than 60 percent of the vote. And they do so almost entirely because of loyal partisan voting.

The hardening of partisan lines has strong geographical implications. That’s because urban areas — and, increasingly, suburban areas — tend to vote Democratic while exurban places tend to vote Republican.

American politics has lots of safe seats — and also ferocious national elections. The two features are related.

The emergence of so many “safe” Congressional seats, a byproduct of the single-seat winner-take-all system, has hugely consequential effects on national politics.

It has spawned a strange duality in American politics. Overwhelmingly, congressional districts and most states are safe for one party or the other. At the same time, in almost every national election the balance of power in Washington is up for grabs.

The result of that state of affairs is that the winning party always sees its majority as threatened; correspondingly, the losing party perpetually views itself as one “wave” election away from unified party government. In a new book, Insecure Majorities: Congress and the Perpetual Campaign, University of Maryland political scientist Frances E. Lee makes a compelling case that this dynamic is a key driver of the partisan polarization in Congress: Both parties are constantly trying to stick it to each other in hopes of winning the next election. Rather than being spread across many competitive districts, the battles are fought in a relative handful of contested seats, with appeals targeting swing voters.

Since partisans of each side are uninterested in compromise, each party’s ability to win depends on casting the other party as too extreme, too terrible, too corrupt, too evil, too un-American — whatever parade of horribles resonates. As a result, “negative partisanship” — partisans hating the other party — is now the most consequential force in American politics.

The past decade has been particularly revealing on this score. In 2009, Democrats held control of the presidency, the Senate, and the House. Republicans decided that their best strategy was to use every possible tool of obstruction, and say as many awful and terrible things about Democrats as they could find a megaphone to say. The plan was to destroy the Democratic Party brand. Politically, it worked. It is now the Republicans who have unified control. Republican voters may not have been enthusiastic about Donald Trump. But to them, Hillary Clinton was far worse.

And Democrats are now where Republicans were in 2009 —- in the minority — facing the same calculus. Since the Republican strategy of pure obstructionism and negation worked, why shouldn’t Democrats use the same strategy?

It’s easy to fall into the “they started it, so now it’s payback time” logic, harder to get out of it. It’s more difficult to take a step back and realize that this escalation is the logical consequence of the zero-sum electoral system we’ve set up.

To see why, look ahead to the likely logic of the 2018 election.

For Democrats to win back the House, they will have to do three things simultaneously in the very limited number of competitive districts they could conceivably win back. They will have to excite their loyal base, convince the few remaining swing voters to support them, and sufficiently demoralize Republican voters that enough of them stay home.

Newt Gingrich 1995
As polarization increased, a scorched-earth brand of politics began to make strategic sense — one often associated with Newt Gingrich.
Richard Ellis / AFP / Getty

Coincidentally, these three things more or less reduce to the same strategy: Make Republicans seem terrible. Raise lots of money to demonize them. Turn Donald Trump into the most evil man in America. And if Republicans are having trouble passing a budget in Congress (or in achieving other legislative goals, even ones some Democrats may be sympathetic with), let them fail spectacularly, and make sure everybody knows it.

Negative campaigning excites core base voters. For them, “Stop The Trump Agenda” is likely to be a great rallying cry.

Meanwhile, swing voters are almost all low-information voters who don’t follow politics very closely. Negative advertising helps them simplify politics into a morality play of good versus evil, which is much easier to grasp than having to understand policy. They’re also probably more likely to vote if they feel their vote matters, so there’s an incentive to persuade them they are voting against a threat to America.

Of course, Republicans will do the exact same thing, hoping to mobilize their own voters by reminding them that a Democratic majority would mean terrible awful things for America.

As a result, we’ll get the same election we’ve grown increasingly numb to: Both parties will descend on the narrow band of swing states and districts, whose citizens’ reward for mattering electorally is that they can’t turn on their televisions or go anywhere without hearing why both the Republican and Democratic candidate are “Wrong for America.”

Meanwhile, in the majority of safe one-party districts and states, members of Congress can simply cruise to victory. And the people voting for them will be voting mostly because of party labels, not anything individual politicians have done or not done (unless it is truly heinous).

A map of Maryland’s congressional districts
Maryland’s congressional districts, as they look today.

These politicians’ only worry is the remote chance that they might get primaried, a lingering threat that keeps them from doing anything that would upset their party’s base voters. If they have any ambition in Congress, they will also remain loyal partisans, scoring points and raising money for the “team” so their party can use the money to reach those few swing voters in those few swing districts. There are few rewards for them to depart from party groupthink to work with the other side to broker deals, and lots of punishments should they try.

What about gerrymandering — the cause of much Democratic handwringing, not to mention lawsuits? It’s exaggerated. No doubt, the Republican redistricting following the 2010 census has given Republicans a notable advantage in translating votes into seats. But as long as we have single-member districts, and as long as Democrats concentrate in cities while Republicans live outside of the cities, any attempt to redraw districts to make them competitive would require awkwardly connecting slices of city to far-flung patches of country in ways that look even stranger and uglier than the current gerrymanders. Such redistricting would also break apart many identifiable communities.

All in all, we’ll get yet another turning up of the “negative partisanship” knob. At what point, as with a speaker that can only handle so much volume, will the system blow?

What might a better system look like in practice?

Our system of single-member plurality-winner congressional districts has accelerated polarization, made most voters irrelevant, and ratcheted up negative partisanship. The big mystery is why we put up with it. Almost all advanced industrial democracies do something different. Why should we accept a system that clearly isn’t working?

Maybe it’s just a failure of imagination: better the devil you know. Or perhaps it’s a version of American exceptionalism: If it’s the American way, it must be the best.

Under a plan endorsed by the group FairVote, Maryland would have two districts — one with five members, the other with three. (Today it has eight districts, with one winner in each.) Multimember districts would make third parties viable, and give Republic
Under FairVote’s plan, Maryland would have two districts, each with four members.

But let’s say we wanted to move to a proportional representation system. How would this work practically in the United States? Worldwide, there are almost as many different systems of proportional representation as there are democracies, and all kinds of different ways in which votes get translated into seats.

Picture how FairVote’s Fair Representation plan would play out in, for example, New York City. (The group has graphic representations on its website of how its plan could work in every state.) Instead of a dozen congressional districts covering varying parts of New York City and Long Island, FairVote’s plan would yield three larger districts.

The current New York City delegation, Staten Island excepted, is overwhelmingly Democratic, and extremely liberal on social issues. This certainly represents the majority of New York voters’ views, but hardly taps into the city’s political diversity.

Under the FairVote system, New York City might send a few moderate Republicans to Congress, plus a more ideologically diverse group of Democrats — and maybe even a few third-party candidates, who would contribute new perspectives and ideas.

Presumably, most candidates elected under this system would be still be from the two major parties, but there’d be space for third parties to grow and develop. And different types of Republicans and Democrats could run against each other, bringing more diversity to the party. A conservative Democrat might make a run in hopes of grabbing a No. 2 position at the polls, even if she had no chance of coming in first.

A Republican voter could vote for a Republican as a first choice, and an acceptable Democrat as a second choice. Why should New York City Republicans not have someone who represents them, even if they make up only 20 percent of the population.

Yes, such a system would weaken both national parties. But individual factions within both major parties could do better.

Nationally, moderate Republicans could brand themselves separately from Freedom Caucus Republicans, and vice versa, and let voters decide how many of each they wanted to send to Congress. (Today they’d be pitted against each other in primaries in which only one faction could win.) Libertarians could finally be their own party, too, without having to compromise on key issues to join one of the two major parties.

Democrats from the Bernie Sanders–Elizabeth Warren wing of the party could brand themselves separately from Democrats from the Clinton wing of the party, and compete alongside each other, again without cannibalizing each other’s votes.

While it would be better to do this nationally, there’s no reason that states couldn’t get the process going on their own. Indeed, last year, then–Maryland State Sen. Jamie Raskin, now a US House member, introduced a “Potomac Compact” with Virginia to end gerrymandering in both states. The compact was intended to be a demonstration project. Maryland is gerrymandered to help Democrats, Virginia is gerrymandered to help Republicans. Should both states approve proportional representation, the bill would empower state redistricting committees to develop multi-member districts. With its 11 seats, Virginia could have two four-member districts and one three-member district. With eight seats, Maryland could have two four-member districts.

Other states are moving toward reform in piecemeal fashion. Last fall, Maine voters approved ranked choice voting for its state representatives, governor, US representatives, and US senators starting in 2018, though opponents have raised legal challenges.

But how would a national proportional-voting system translate into political behavior in Congress? What would happen if — because of those pesky third (or fourth) parties — no party had a majority? How could Congress function?

It would function just like most democratic legislatures around the world. After the election, parties would form governing coalitions based on how many seats they won. The nice thing about such coalitions is that they can be flexible, and even fluid, depending on the issues, which would create opportunities for different factions in Congress to work out deals on different issues. For example, if there were more factions willing to play ball on health care reform, it’s possible a deal could get worked out. Or if most factions wanted to marginalize extremists, they could do that too.

If this seems unrealistic in the American context, recall that this is sort of how Congress used to work. From the 1930s through the 1980s, both parties were more like loose coalitions of different factions; depending on the issue, different coalitions formed. Notably, key civil rights bills in the 1960s passed with votes from both Democrats and liberal Republicans.

Particularly in the 1970s, when both parties had overlapping coalitions and party leaders were weak by today’s standards, Congress gave committees both the staff resources and autonomy to solve public problems and oversee the executive branch. As a result, Congress was at the height of its institutional power.

But as both parties became more homogeneous, and delegated more power to their leadership in Congress, coalitions became much less fluid, and the committees have atrophied. Congress ceased to be an institution and instead became a partisan battleground, where senators and Congress members were Republicans and Democrats first, and members of Congress second. Everything became campaigning.

Our electoral rules stink. Let’s change them.

Here’s the bottom line. Our single-member, winner-take-all approach to elections might have worked okay during a period of low polarization, when parties were overlapping coalitions, when more places had genuine two-party competition, and when voters were sometimes willing to support candidates from the other party.

But those things are in the past now. Our electoral rules are now gasoline for the current conflagration of partisan polarization. Because the polarization is primarily ethnocultural and therefore geographical, congressional districts and most states are safe for one party or the other. But in almost every national election, the balance of power in Washington is up for grabs.

Map showing how under the FairVote system, Ohio would have four House districts, with three to five seats each. A candidate could win a seat with as little as 17.5 percent of the vote.
Under the FairVote system, Ohio, too, would be divided into large multimember districts, spanning urban and rural areas.

That means that both parties are always drawn to the siren song of unified party control, at which point they can finally enact their agenda — an end that justifies almost any means. And because of the two-party nature of the competition, Democrats and Republicans each have only one enemy: each other. They win more by tearing each other down than by running on their achievements — and they don’t have to worry some third party will sneak in and steal their voters.

But they are fighting over less and less contested territory, which makes the remaining fights even more bitter and consequential. And it reduces most voters to the status of bystanders.

Sure, committed advocates could physically move to Iowa (or the few other closely contested places) to make their votes count. But it would be far better if the entire nation shifted to an electoral system where everyone’s vote matters, regardless of where they live. To a system in which the incentives did not push political parties into zero-sum trench warfare, but toward compromise and coordination that would solve pressing public problems.

In short: Don’t move to the Midwest, young hipster. Become an advocate for proportional voting. It wouldn’t just be young people, or just Democrats, who would benefit from the reform.

Lee Drutman, a regular contributor to Polyarchy, a Vox blog, is a senior fellow in the political reform program at New America. He is the author of The Business of America is Lobbying: How Corporations Became Politicized and Politics Became Corporate (2015).

The Big Idea is Vox’s home for smart discussion of the most important issues and ideas in politics, science, and culture — typically by outside contributors. If you have an idea for a piece, pitch us at