As Part 2 in this series explained, the US Supreme Court could adopt a test to protect against intentional partisan gerrymandering. Such a test could prevent egregious power grabs like Wisconsin’s Republican-drawn district maps that enabled Republicans to win more than 60 percent of seats in the state house despite winning less than 49 percent of the vote.
But unintentional unfairness is the scaly underbelly of single-winner districts. For example, Washington state uses an independent redistricting commission to eliminate intentional bias, but in 2016 Republicans won 51 percent of seats in the state senate even though Democrats won 54 percent of the votes.
Partisan districtors can use district lines to “pack” and “crack” voters of one party. But it turns out voters unwittingly “pack” themselves when they move to places where they are surrounded by like-minded people. If some voters have “packed” themselves into certain geographical areas, the most disinterested computer algorithm, seeking to make compact, un-salamanderish districts, will end up disproportionately diluting those voters’ votes. Other voters’ votes will have more power to elect a representative, and the legislature will unfairly skew towards the geographically more dispersed voters.
Urban Democrats have “packed” themselves
In the United States today, left-leaning voters have packed themselves into dense urban areas. Researchers call it the “Big Sort” and “natural geographic sorting.” Journalists recognize the “two Americas.” Part of why the Supreme Court has been reticent to strike down gerrymandering is the difficulty in distinguishing intentional partisan manipulation from the “natural packing effect.”
Some research concludes that unintentional self-sorting plays a bigger role in the “persistent pro-Republican bias” in districts than intentional gerrymandering, and even when maps are drawn by courts or commissions, Republicans will still win more seats than votes. The Republican National Committee’s brief to the Supreme Court in the upcoming gerrymandering case dedicates an entire section to this point, titled: “Research confirms that the geographic distribution of Republican and Democratic Voters result in more Republican Seats irrespective of partisan gerrymandering.” It points to computer modeling showing that, in single-winner districts, geographically concentrated voters can’t win as many seats as geographically dispersed voters, even if they outnumber them.
The researchers say “as the ratio of seats to voters in a legislative body increases, it becomes more difficult for the political party whose voters are geographically concentrated to maintain statewide vote parity.” In other words, smaller single-member districts make it difficult for geographically concentrated (in dense urban areas) Democratic voters to win a fair number of seats. A Democratic-leaning state will elect a Republican-leaning legislature from single-member districts even if a non-partisan body draws the lines. (See: Washington state senate.)
Compact districts make rural voters more powerful than urban voters
Redistricting processes, including California’s well-designed independent commission, often include an explicit goal of drawing compact districts, to avoid the unsightly shapes that signal intentional gerrymandering. Sure, independent commissions or computer algorithms can draw compact districts. But, because of the “natural sorting” explained above, those compact districts will disempower voters; they will be neither competitive nor representative.
Here’s a simple thought experiment modified from the one described in this article: a state has 56 Democratic voters and 44 Republicans divided into 10 single-member districts. The state’s four urban districts vote 8-to-2 for Democratic candidates and the six suburban and rural district vote 6-to-4 for Republican candidates. Democrats have a clear majority of voters but only 40 percent of the seats—a legislature that doesn’t represent the voters. No nefarious partisan shenanigans needed—just single-member districts’ inherent flaws. Washington voters aren’t as neatly divided as this thought experiment, but the state senate shows this principle in action. Sixty percent of Republican senators won with less than 60 percent of the vote, but only 25 percent of Democratic senators did. Half of Democratic senators won with a landslide more than 70 percent of the vote. More concentrated Democratic districts allowed Republicans to win more seats than they won votes, based solely on geography and not intentional gerrymandering.
Original Sightline Institute graphic, available under our free use policy.
Safe districts make political parties and primary voters more powerful than general voters
In the example above, none of the districts are competitive. Whichever candidate wins, the party primary can just coast to victory in the general. When an incumbent is running in the primary he is almost guaranteed to win—in Oregon more than 90 percent and in Washington more than 85 percent of state house incumbents face no primary challenger. In a safe district, members of the weaker political party knows they have no chance to win, so they might not even bother running a candidate, leaving general election voters literally without any power to choose who wins. In close to 40 percent of races in Oregon and Washington, one of the two major parties doesn’t even bother running a candidate.
An independent commission could instead aim to draw competitive districts where each party has a shot at winning and each candidate has to work to connect with voters. To do that, the line-drawers would have to slice and dice those cities into irregular, gerrymandered districts that clumsily concatenate a bit of the city and a chunk of suburbia. Pizza slices radiating from downtown Portland or Boise, for example, or ribbons that cross Washington from the Pacific to the Idaho border. Independent commissions are loathe to draw districts that might look at a glance to be gerrymandered, so they end up drawing districts that are just as safe for one party as do partisan line-drawers.
Single-winner districts minimize voters within and outside the two major parties
Measuring how many votes and how many seats each party wins doesn’t account for the voters who might prefer a different candidate within their own party or another party altogether. For example, in 2016, Libertarian candidates won 3 percent of the vote in Washington, despite the fact that single-member districts systematically exclude third parties from winning seats. Likely, more voters truly preferred a Libertarian or other third-party candidate but didn’t want to throw their vote away on a candidate doomed to lose. All those voters have less power to elect a representative they like than do voters who truly prefer a major party candidate.
Even within a major party, voters may differ. For example, in 2016 56 percent of voters in Washington congressional District 7 voted for Democrat Pramila Jayapal, while 44 percent voted for Democrat Brady Piñero Walkinshaw. The rough two-party analysis says 100 percent of those voters got a representative from their preferred party, but clearly one wing of the party won representation while voters from another wing may not feel as represented.
One person, one vote
The US Supreme Court has long upheld the principle of “one person, one vote.” This principle means districts must include roughly the same number people—otherwise, voters in a less populous district would have more power than voters in a more populous district. The Supreme Court has held that racial gerrymandering—packing African-Americans into a single district—illegally discriminates based on race by giving white voters more power than the packed African-American voters. But the Court is not likely to weigh in on the idea that single-member districts give rural voters and primary voters more power than urban voters and general election voters, but the same principle applies. Every voter should have equal power to elect someone to represent her in the legislature.
Multi-member districts make all voters count equally
In a district where three candidates can win seats, gerrymandering can’t disempower voters. In a district that would be completely safe for one party in a single-member system—say it leans 70 percent Democrat—conservative voters would be able to elect one of three representatives. And even the Democratic incumbent couldn’t just skate by without voters having a say because two left-leaning seats would be up for grabs, opening the race to a different kind of Democrat or even to a Social Democrat or a Green Party representative. Voters of all stripes would have equal opportunity to make their vote count in electing a representative of their choice.
Multi-member districts give all voters a chance to elect a representative, no matter their views or their zipcode.
For example, currently, each of Washington’s 49 districts elect three state officials (a senator and two state representatives) in the three separate races, and often all three are of the same party, and even the same wing of that party. A voter in that district who prefers a different party or a different wing of the party has no one representing her views in the state legislature, even though three people represent her district. If the voter could instead elect three officials at once, she could make her voice heard as loudly as every other voter. The result in a left-leaning district might be a switch from three similar Democrats to one progressive Democrat, one standard Democrat, and one moderate Republican.
In the 2016 race for congressional District 7, voters were forced to vote either for Pramila Jayapal or Brady Pinero Walkinshaw. But with the federal Fair Representation Act’s multi-member districts, progressive voters could have ranked Jayapal first and Walkinshaw second, while conservative voters could have voted for someone who better matched their views.
Multi-member districts give all voters a chance to elect a representative, no matter their views or their zipcode.
Greg Cusack
Again, my compliments, Kristin, for your clear exposition of the advantages of multi-member districts. You have convinced me!
I have a question for you about a related issue which has a lot to do with the “mechanics” of actually implementing multi-member districts: the nature of “representation” in a huge republic.
As you know, at the time of the forming of our Constitution many political theorists doubted that a republic could survive covering a large territory, a concern shared by several of the Founders as well. While part of this certainly had to do with the limitations of communication in the 18th century, a related issue had to do with a republic’s success deemed more likely when the interests of the people were more homogeneous and the people were more able to watch over the actions of their representatives.
The original language of the Constitution attempted to ensure some measure of accessibility by the people to their representative by saying that the upper limit for representation was one person for every 30,000 citizens. Today, because the size of the House of Representatives is limited by law to 435, the relevant figure is one for every 751,000 citizens.
Clearly, it is impossible for any one human being to somehow “represent” such a huge segment of the population. Personally, I am convinced that this has a lot to do with what has happened to the behavior of the House in recent decades.
However, my question to you is this: With multi-member districts in mind, would each of the existing House districts be converted into multi-member districts, thus expanding the number of House members and, theoretically at least, restoring some reality to “representation” or, alternatively, would the number of existing House districts be diminished in order that the total membership of the House remain at its currently fixed level? (Of course, for some states like Alaska and Delaware, this is a moot point as they already have only one at-large district.)
Thank you!
Kristin Eberhard
It could be either. Oregon could keep its five (probably soon six) federal representatives and elect them in a state-wide district.
Or lets take the example of the Oregon House–you could keep it the same size (60) and split into 12 five-member districts.
Or, my preference, combine the state house and senate into a single unicameral body with the same total number of Oregon state legislators (90) and elect 30 of them from single-member districts the same size as the current senate districts, and the other 60 from ten-member districts the same size as the soon-to-be six federal representative districts.
Bob Nichols
Good point. It seems in fact that America is too large and polyglot to be effectively governed as a single nation state any longer. It could ideally be broken up into 10 or so smaller, more manageable nations, each with a much higher degree of economic and cultural cohesion than the existing meganation
Dave Corkran
The multi member district system may make passing legislation more difficult. It might mean that many different parties would form, as is the case in some European countries where a single legislature is the dominant political authority. In those countries alliances of different parties are sometimes necessary to gain a majority in a parliament. To achieve alliances, compromises are necessary. But the three part system of American governments (legislative, executive, judicial) means that a party has to control all three branches before it can be sure of making a law that will stick. The need to control three branches may be a reason for the American two party system and the winner take all voting process. Going to a multiple member district voting system might give more people equal representation. But it might also paralyze American governments by creating many more political parties. American governments were set up to protect against the tyranny of the majority. Maybe American political history revolves around the effort to overcome this suspicion of majorities in order to achieve a government that works. The multi member voting district could fracture the body politic into so many factions that reaching agreement among groups is impossible. It is much easier to reach compromise with two groups than with eight or ten. Perhaps our author needs to think through her proposals from some different perspectives.
Kristin Eberhard
The US system is pretty paralyzed and fractured right now. It see-saws back and forth between two parties with razor-thin margins and no willingness to compromise, making it impossible for them to reach agreements, increasingly unlikely they can get anything done, and hard to make make anything stick.
In contrast, PR countries craft solutions that a majority coalition likes, and those solutions stick. We don’t have to speculate about whether PR countries break down, because nearly every developed country in the world uses PR, and it works. http://www.sightline.org/2017/05/18/sightlines-guide-to-methods-for-electing-legislative-bodies/#proportional-solution-all-interests-have-a-voice-in-shaping-consensual-policies
Our system governs with a strong but unsteady hand, veering back and forth between two uncompromising parties. PR countries govern with a gentler, but steady hand, with many parties that work together to get things done.
http://www.sightline.org/2017/05/18/sightlines-guide-to-methods-for-electing-legislative-bodies/#proportional-solution-voters-learn-about-and-legislators-craft-innovative-durable-policy-solutions