The Washington PostDemocracy Dies in Darkness

Here’s a different way to fix gerrymandering

Even independent redistricting commissions are limited by the fact that Democrats cluster in cities while Republicans are widely dispersed. But there’s a solution.

Demonstrators at a rally at the Supreme Court as it hears the gerrymandering cases Lamone v. Benisek and Rucho v. Common Cause, March 26, 2019. (Evelyn Hockstein/For The Washington Post, file)
Placeholder while article actions load

As state lawmakers start drawing state legislative and congressional districts, Americans can be confident about this: Many districts will be drawn in bizarre shapes that give one party the advantage. While new technologies have made gerrymandering much easier, citizens have helped by clustering in neighborhoods with others who are politically like-minded, making it easier to see how any region is likely to vote.

But some local and state elections employ “multi-member districts.” These districts have a larger population than single-member districts and elect more than one representative. That represents more views and promotes competition. Here’s how they work.

Americans have sorted themselves in ways that make it harder to draw “fair” maps

Over the past few decades, Americans have increasingly sorted themselves into neighborhoods of the politically like-minded, scholars find. This “partisan sorting,” as it’s called, reinforces longstanding patterns of housing segregation by race and income. With fewer and fewer politically diverse communities, state legislators find it easier to pack their opponents’ voters into a single district – called “packing” – or to dilute their votes by dispersing such communities into many different districts – called “cracking.” Either way, those voters are marginalized. Political scientists say many of their ballots are “wasted” – delivered in ways that don’t affect the outcome, whether poured into huge margins in packed districts or ineffectual votes in cracked districts. Meanwhile, incumbents are so secure that they only fear competition in low-turnout primaries.

Many reformers advocate for nonpartisan, independent redistricting commissions that will draw “fair” maps. But even when district boundaries are drawn in reasonable ways, Democrats are concentrated in large, urban areas, while Republicans are spread across suburbs and rural areas. As a result, GOP candidates win with smaller margins -- wasting fewer votes -- and claim more seats than their share of the state vote would suggest.

Some election experts recommend multi-member districts, which counteract partisan sorting in a variety of ways. Increasing each district’s size makes both packing and cracking harder, since each district is larger than even the most populous partisan cluster. Fewer districts mean fewer boundaries to manipulate. Larger districts offer more coalition possibilities to attract strong, well-funded candidates from both parties. Together, these changes encourage both incumbents and challengers to attend to a wider range of voters.

Here's how to fix partisan gerrymandering, now that Congress has kicked it back to the states.

The mechanics of multi-member districts

Multi-member districts are as old as the Republic and are still used in municipal and state elections. In the 10 states that use multi-member districts for the lower house, procedures for determining the winner depend on how primaries are structured, votes are tallied, and how many candidates are nominated. In a hypothetical three-member district, each voter typically casts three ballots; rules vary on whether a voter can abstain from some or allocate all three votes to the same candidate.

In our forthcoming research on possible multi-member districts in Pennsylvania, we used a computer algorithm to repeatedly and randomly divide Pennsylvania into six congressional districts containing approximately 2.1 million people each, three times the size of the eighteen districts the state has now. In the tens of thousands of valid maps that resulted, each of these larger districts closely matched the state on key demographic categories, such as age, education, income levels, race, urban or rural. Most important, they also contained fewer lopsided partisan blocs, increasing the chance of competitive elections and reducing the number of wasted votes.

A key provision for making multi-member districts work most effectively is Ranked Choice Voting (RCV), in which voters order candidates by how much they prefer them — first choice, second choice and so on – as has just happened in New York City’s mayoral primary. The technical details can be complex and hotly debated by experts. In the simplest method for a hypothetical three-member district, the candidate with the most first-choice votes gets the first seat. If this candidate gets any more votes beyond the 33.3 percent needed to be elected in a three-member district, the computer tabulating those ballots would examine the share of second-choice votes among people who voted for the top candidate, and would allocate those “additional” votes to those other candidates. The votes would be tallied again, and the candidate with the most votes would win the second seat. Once again, additional ballots would be reallocated, and the remaining votes would determine the winner of the third seat.

Redistricting might gain Republicans a few seats in Congress. Their real gains will be in state legislatures.

The challenge of Black majority-minority districts and multi-member constituencies

Since the Voting Rights Act of 1965, state legislators have been drawing majority-minority districts, thus increasing the number of Black lawmakers in both Congress and state legislatures. But the price has been steep: Black voters have less influence over policy. Packed into majority-minority districts, Black voters are easily ignored by elected officials who never face a significant bloc of them.

As Black turnout increases, however, majority-minority districts may no longer be necessary to elect Black candidates. In our forthcoming study, we found that of the districts in which at least 37 percent of the eligible voters were Black, all but two elected a House member who affiliated with the Congressional Black Caucus in the 115th Congress (2017-18). Recent scholarship on Ranked Choice Voting (RCV) and minority representation suggests that multi-member districts with some form of ranking built into the ballot could help Black and Latino communities elect their top choices.

To make this possible, Congress would have to repeal a ban on multi-member districts enacted in 1967, when Southern legislatures were using at-large elections without ordered preferences to swamp minority candidates. With RCV, however, candidates who are the top choice of a significant minority of voters would prevail over the majority’s less preferred candidates.

Multi-member districts with ranked-choice voting do not guarantee that each party will hold the proportion of seats that matches the proportion it wins of the vote. But the system would lessen the likelihood of wasted votes. Large majorities would secure their first choice, and perhaps their second; large minorities would see at least one of their candidates elected as ballots were reallocated. Seats would be more likely to be competitive, and if voters’ second and third choices matter, candidates would have more reason to reach out to voters now largely ignored in winner-take-all elections. We think citizens in both parties would be better off.

Don’t miss any of TMC’s smart analysis! Sign up for our newsletter.

Christopher S. Fowler (@csfowler13) is associate professor of geography at Pennsylvania State University and has published extensively on the consequences of population clustering in a wide range of communities with particular emphasis on racially segregated and immigrant communities.

Linda L. Fowler is professor of government and Frank J. Reagan Chair in Policy Studies, Emerita, at Dartmouth College, and has written widely on congressional elections and the recruitment of candidates.