Nonrepresentational Line-Drawing and the Universal Representational Imperative: Why Judges Should Replace Gerrymandering with Proportional Representation

People complain that gerrymandering ruins redistricting.[1] But, in the real world, redistricting is gerrymandering. If you think elected officials redrawing legislative districts do not know exactly who lives in those districts in terms of party, race, ethnicity, and potential challengers, you may be too innocent to be let out of the house by yourself. Whatever their other specialties, all incumbents are, at least decennially, experts on the political demography of their own districts,[2] and the necessary logic of redistricting as an incumbent is to choose your voters on the map before they choose you on the ballot.

Computer technology has exacerbated this native political tradition in the United States by enlarging the “bag of tricks” available to electoral cartographers. Many districts have come to look exceedingly strange as a consequence.[3] This is not to say that politicians have ever been known to draw rectangular, circular, or triangular districts before—the word “gerrymander” itself dates back to 1812,[4] and the tradition of gerrymandering is as American as apple pie or, perhaps more precisely, as American as slicing an apple pie in such a way as to feed your friends and family while starving your enemies.[5] What is new today is the way that computer-enabled map-drawing pulls the visual appearance of districts away from recognizably figurative and “representational” geographic forms, the equivalent of a still life by Matisse or Braque, to “nonrepresentational” forms that share a much greater affinity with the abstract expressionism of an artist like Kandinski.

Consider the amazing new Third Congressional District drawn in Maryland, which sweeps in several precincts that I represent in the Maryland Senate. To be sure, the Third District has always covered a lot of political ground and, probably for that reason, has produced the state’s last three United States Senators—former U.S. Senator Paul Sarbanes and the present incumbents Senators Barbara Mikulski and Ben Cardin, all of whom represented the Third District in the U.S. House before getting elected to the Senate.[6] But the Third District formerly corresponded roughly to recognizable geographic and political communities; today it links bits and pieces of communities strewn from Baltimore City and Baltimore County all the way to Annapolis, Howard County, and East Silver Spring in Montgomery County. To the voters, the composition of the district makes very little sense, despite the fact that it has an exceptionally fine Congressman today in John Sarbanes.

As a result of this kind of partisan line-drawing and high-tech community‑fracturing, the public is displeased, baffled, and hungry for a change, but no one knows how we might evolve beyond gerrymandering.[7] If you stare deeply into the new post-figurative redistricting maps that have emerged all over America and use them like political Rorschach tests, what do you see?[8] Most people see only the Machiavellian designs of incumbent politicians, and the media routinely reinforces this frame of political cynicism. What I see, a bit more hopefully, is the end of the road for winner-take-all elections in single‑member districts, the system that the states are still forced, by federal law, to use for choosing members of the U.S. House of Representatives.[9]

I say this because the oddly-shaped electoral maps in many states only make plain graphically what is going on politically in almost all states: the triumph of incumbent and partisan self-entrenchment by the dominant parties in state legislatures. The controlling imperatives of reelecting incumbents and maximizing the number of seats claimed by the top-dog party squeeze out all other interests. Thus, the goals of incumbent reelection and partisan advantage subordinate all other concerns: the “traditional” redistricting interests of the states, including maintaining communities of interest, compactness, and meaningful contiguity; the goal of empowering racial and ethnic minority groups in legislative chambers as required by the Voting Rights Act;[10] and, of course, more aspirational and visionary purposes, such as getting more women elected to office or making certain that all substantial political viewpoints in a state are represented. The primacy of incumbency and partisan advantage also means that legislative districts become more safely partisan over time—the Democratic ones more Democratic, the Republican ones more Republican—thereby thwarting the virtues of competition and steadily eroding voter interest and participation.[11] Meanwhile, as courts are increasingly drawn into the redistricting process by parties and candidates grasping onto the numerous available legal pegs,[12] the judiciary itself inescapably becomes implicated in a process that consecrates the desires of ruling parties and incumbent politicians over every other constitutional or policy interest in the field.

Recent events in Maryland exemplify the partisan dynamics. My Republican colleagues in the Maryland Senate complained about the majority Democratic Party redrawing congressional district lines to make Democrats competitive in the new Sixth Congressional District, which merges Western Maryland with more of the Washington suburbs of Montgomery County than ever before.[13] The new district gives Democrats a narrow registration majority and a decent shot at beating incumbent Republican Congressman Roscoe Bartlett, thereby capturing a seventh seat out of the eight in the state.[14] The Republican minority naturally wanted the seat to remain safely in Republican hands and cried foul.[15] But the Maryland Republicans’ righteous indignation is shared by minority Democrats in many other states right now.[16] Take North Carolina, which then-Senator Obama won in the 2008 general election.[17] There, the Republican-controlled legislature carved up the state to achieve ten supermajority‑Republican districts and three supermajority-Democratic districts.[18] Or consider Ohio, another state that Obama carried in 2008, whose Republican-controlled legislature drew twelve majority-Republican districts, leaving huge crowds of Democrats packed into just four Democratic districts.[19] And so, across the country, the race between the parties to nail down every last congressional seat accelerates, with controlling parties seeking to pack as many rivals as they can into a handful of districts while locking in a ten-year advantage in all the others. No state legislative caucus will “unilaterally disarm,” as the politicians say in the cloakroom, which means that the people driving the process are thinking as strategic partisans rather than as governors in pursuit of the common good.

This tawdry process demeans democracy, but the swerving district lines that are so catchy to the eye and so alluring to editorial writers are actually the least of the current system’s offenses. The real problem is the broad failure of political representation produced by all of our single-member districts. After all, congressional districts should be judged not as works of fine art, which they never were and never will be, but on their ability to produce effective political representation, which is the essential promise of representative democracy. Independent redistricting commissions focused on aesthetics may succeed in improving appearances, but, from the standpoint of parties and groups stuck in the permanently losing minority position at the statewide level, even tidy‑looking districts feel sharply lopsided and inequitable.[20]

Imagine a state with ten congressional districts where the Orange Party has 60% of the state’s voters spread evenly across the state and the Purple Party has 40%. In most democratic countries, which conduct elections based on proportional representation,[21] the Orange Party would end up with six seats and the Purple Party would get four. Seems fair, right? Each party gets representation proportionate to its political support. But, in America, the Orange Party will have not only the motivation but also the technological capability and the legal authority to draw eight, nine, or even ten majority-Orange districts and thus run the table on their opponents. The problem with such a system is obvious: 40% of the people are essentially unrepresented. Surely it’s right for the majority to rule, but surely it’s wrong to deny the minority a voice and a place at the table. Even when rendered in the form of aesthetically appealing districts, winner‑take-all politics is a pretty ugly thing. And even if an independent districting commission draws contorted districts to advance partisan competitiveness, which is indisputably a significant value, a row of districts split 51%-49% actually leaves many more people feeling politically dissatisfied and marginalized than districts that are drawn 65%-35%. In the system of single‑member districts, when we increase competiveness, we reduce representativeness; conversely, when we increase representativeness, we reduce competitiveness.

These problems are intractable without a paradigm shift. It is time to replace the winner-take-all, single-member-district mentality with a system that produces what John Stuart Mill considered full representation, in which all citizens are seen and heard through their chosen representatives.[22] The idea is that nearly everyone, if not actually everyone, will help to elect representatives who mirror their views and sentiments. This is what I think of as the “universal representational imperative.”

Consider the congressional map, produced at my request by the electoral think tank FairVote, which I circulated to colleagues in Annapolis. The map creates two multimember “super-districts” in Maryland.[23] Voters would elect congressional representatives using a fair voting system such as choice voting, in which voters rank all their choices and candidates win when their vote totals hit the appropriate numerical threshold,[24] or cumulative voting, in which voters have the same number of votes as there are seats and distribute them as they see fit.[25] In District B, where there are five representatives, the threshold of victory for both systems would be 17%, which means 17% of the voters could target their votes on one candidate and get him or her elected to Congress. A system like this gives voice to minorities, however defined, and would bolster our efforts to diversify our congressional delegations along the lines of political party, ideology, race, and ethnicity.[26]

Proportional regimes also promote gender diversity and should be considered feminist electoral systems. In Maryland, a leading progressive state, we have only one woman—Congresswoman Donna Edwards—out of eight Members in our House delegation.[27] This imbalance would be extremely unlikely with multimember districts, where people get to vote for several candidates and diverse, gender-balanced slates usually form in order to mirror and persuade the electorate.[28] All over the world, women candidates emerge and thrive in systems of proportional representation—not only because of the likelihood of slate formation, but also because such systems do not operate on the basis of the kind of ad hominem negative politics that seems to be an intractable feature of many single‑member districts and, in my experience, tends to alienate many potential female candidates.[29]

The best argument for preserving the status quo single‑member district system, of course, is that it ensures that specific territorial political communities are coherently served and represented.[30] But this argument loses much of its force when congressional districts are composed of upwards of 720,000 people and the district perimeters freely cross county, municipal, historical, and natural boundaries like the coloring designs of a preschooler who cannot stay within the lines. When we look at maps of today’s districts, such as Maryland’s First Congressional District or Third Congressional District, it seems hard to argue that single-member districts facilitate the creation or preservation of meaningful political, geographic, or historical communities. Two multimember districts in Maryland would guarantee some basic geographic distribution while essentially permitting the voters to “district” themselves according to the criteria and characteristics they find most significant in each election. Although there is mounting displeasure with the redistricting process, which combines bare‑knuckled politics with the aesthetics of abstract expressionism, a change in voting systems will be hard because state-elected officials are profoundly risk‑averse when it comes to altering the rules of the game, and because no state wants to jump first and practice unilateral disarmament in the race to achieve political control through the art of redistricting.

But there is an even more fundamental and daunting threshold obstacle to changing the single-member districts: Even if a state wants to act, there is nothing it can do absent a change in federal law because Congress has, since 1967, required the states to use single-member districts in all congressional races.[31] To vindicate the consent of the governed in the way that we need to advance it today, Congress would have to change the law to give states the power to set up independent redistricting commissions to develop multimember district elections. In other words, the members of Congress most deeply invested in maintaining the current regime—incumbents of both parties—are the ones who will have to decide to break from the current orthodoxy.

This congressional self-interest means that the judiciary must play a key role in initiating proportional voting alternatives that are superior to single‑member districts in securing basic representational values. Professor Steven Mulroy points out that several courts have already adopted proportional-voting systems as remedies to Voting Rights Act violations at the local level.[32] These courts recognized the impossibility of shoehorning all of the representational and democratic aspirations of voters and political communities into the pinched confines of single-member districts. The judges in these cases were willing to experiment with the forms of electoral democracy in order to enhance its essential purposes.

 While it may seem daringly original or avant-garde for judges to order multidistrict proportional-voting remedies in voting-rights cases, the alternative is to remain passively complicit with the aggressive self-entrenchment of incumbent officials and majority parties that is the baseline structure of every redistricting map in America. But why should judges take as a baseline for democratic representation a particular design that has proven itself so seriously flawed when superior alternatives are readily available?

 As we look for ways to transcend the glaring defects of single-member districts and the redistricting process, the dynamics of political self-interest make it likely that courts, along with third parties, independent voters, and civic reform movements like the League of Women Voters and Common Cause, will have to become the driving catalysts for change. If these outside actors fail to shock the system, the maps will grow ever more fanciful, the public will grow ever more disenchanted, and the processes of representation will remain stuck, antiquated, and profoundly demoralizing.

[1].        See Aaron Brooks, The Court’s Missed Opportunity To Draw the Line on Partisan Gerrymandering: LULAC v. Perry, 126 S. Ct. 2594 (2006), 30 Harv. J.L. & Pub. Pol’y 781 (2007); Laughlin McDonald, The Looming 2010 Census: A Proposed Judicially Manageable Standard and Other Reform Options for Partisan Gerrymandering, 46 Harv. J. on Legis. 243 (2009).

[2].        See Monica Davey, New Faces, New Maps, New Battles, N.Y. Times (July 14, 2011),; Nate Silver, In Politics, Demographics Are Not Destiny, N.Y. Times FiveThirtyEight Blog (Mar. 1, 2011, 6:49 PM),

[3].        See, e.g., Micah Altman & Michael McDonald, The Promise and Perils of Computers in Redistricting, 5 Duke J. Const. L. & Pub. Pol’y 69 (2010); Josh Goodman, Redistricting: Should Computers Draw the Lines?, Governing (July 26, 2010),

[4].        See The Gerrymander, Libr. Congress,
trr113.html (last visited Nov. 17, 2011).

[5].        For a history of redistricting in the United States, see Reapportionment Politics: The History of Redistricting in the 50 States (Leroy Hardy et al. eds., 1981).

[6].        For present and former U.S. Senators and Representatives from Maryland’s Third District, see Maryland’s 3rd Congressional District, Open Congress, (last visited May 2, 2012).

[7].        To gauge public opinion, see, for example, Editorial, Gerrymandering, Pure and Corrupt, N.Y. Times (Nov. 11, 2009),
opinion/12thu1.html; and What Others Are Saying, Gerrymandering, http:// (last visited Nov. 15, 2011), which collects nonpartisan sources and states that more than 71% of respondents to a survey disapproved of the current Illinois system of drawing legislative districts. Id.

[8].        See Aaron Kase, Census Puts Gerrymandering in the Spotlight, Phila. Weekly (Mar. 16, 2010),

[9].        2 U.S.C. § 2(c) (2006).

[10].      See Voting Rights Act of 1965, 42 U.S.C. § 1973 (2006).

[11].       For historical reference on reelection rates from 1964 to 2010 in the U.S. House of Representatives and U.S. Senate, see Reelection Rates Over the Years,, (last visited Apr. 21, 2012), which notes that reelection rates since 1982 have remained between 85 and 90% for House incumbents and around 80% or above for Senate incumbents, with a few outlying years. See also U.S. Census Bureau, Statistical Abstract of the United States: Elections 257 (2008) (demonstrating a general decline in voter-turnout percentages in both presidential and congressional elections from 1932 to 2006, punctuated by occasional spikes upward).

[12].      Parties, candidates, and voters with an ax to grind can allege violations of the constitutional one person, one vote command under Wesberry v. Sanders, 376 U.S. 1, 18 (1964), and Reynolds v. Sims, 377 U.S. 533, 586 (1964); Section 2 (or, in some cases, Section 5) of the Voting Rights Act of 1965, 42 U.S.C. § 1973 (2006); the racialized double standard disfavoring majority-African American and Hispanic districts found in Shaw v. Reno, 509 U.S. 630, 657 (1993), and Miller v. Johnson, 515 U.S. 900, 920 (1995); as well as an abundant assortment of relevant state laws and constitutional provisions. For a discussion of why Shaw represents a double standard, see Jamin B. Raskin, The Supreme Court’s Racial Double Standard in Redistricting: Unequal Protection in Politics and the Scholarship That Defends It, 14 J.L. & Pol. 591, 622 (1998).

[13].       See Andrew Schotz, Shank, Myers Show Interest in 6th Congressional District Race, Herald-Mail (Hagerstown, Md.) (Oct. 22, 2011), http://articles.herald
-maryland-s-sixth-district; Redistricting Public Hearings Begin in Maryland, CBS Balt. (July 25, 2011),

[14].      See John Fritze & Annie Linskey, New Map, New Politics in Western Md., Balt. Sun (Oct. 22, 2011),

[15].       Id.

[16].      For a survey of that opinion, see Josh Kraushaar, Democrats’ Losses Could Grow in 2012, Nat’l J. (Nov. 10, 2010),
-the-grain/democrats-losses-could-grow-in-2012-20101110; Robert McCartney, A Pox on Both Parties for Partisan Redistricting, Wash. Post (Sept. 7, 2011),
-redistricting/2011/09/07/gIQAveDDAK_story.html; Michael D. Shear, Winner in Redistricting Is Still To Be Determined, N.Y. Times Caucus Blog (Nov. 18, 2011, 1:46 PM),
-advantage-goes-to/#more-180063. The massive Republican investment in the 2010 state legislative elections, which answered the landslide victory of Democrats in 2008 and produced twenty-six Republican-controlled legislatures, is paying off handsomely now. See Jeremy P. Jacobs, Devastation: GOP Picks Up 680 State Leg. Seats, Nat’l J. (Nov. 4, 2010),
2010/11/devastation-gop.php; Republicans Exceed Expectations in 2010 State Legislative Elections, Nat’l Conf. St. Legislatures (Nov. 3, 2010),

[17].       Katharine Q. Seelye, Obama Wins North Carolina, N.Y. Times Caucus Blog (Nov. 6, 2008, 12:38 PM),

[18].      Rob Richie & Jais Mehaji, Mitigating the Pernicious Effects of Gerrymandering in North Carolina: The Super‑District Alternative, FairVote (July 29, 2011),…

[19].      Sheahan Virgin & Lindsey Needham, No More Gerrymanders: Ohio’s GOP-centric Plan Versus the FairVote Super District Alternative, FairVote (Oct. 13, 2011),

[20].     For a sampling of the inequity that such a single-district system produces, see Greg D. Adams, Legislative Effects of Single-Member vs. Multi-Member Districts, 40 Amer. J. Pol. Sci. 129 (1996); and Lani Guinier, The Representation of Minority Interests: The Question of Single-Member Districts, 14 Cardozo L. Rev. 1135 (1993).

[21].      See Rob Richie, Fair Representation News: Proportional Voting in American Cities and Around the World, Huffington Post (Dec. 19, 2009, 9:46 AM), http://www (emphasizing that most of the democratic world has moved away from winner‑take‑all elections and that the United States should follow suit).

[22].      John Stuart Mill, Considerations on Representative Government 133 (1861).

[23].      Maryland Gerrymander Triggers Support for Fair Voting, FairVote,… (last visited Nov. 15, 2011). The map creates a five-member district and a three‑member district, rather than two four-member districts, because a party collecting a majority of the vote at somewhere between 50% and 60% ends up with only two seats in the four-member district. Thus, if Party A collects 59% of the vote and Party B 41% of the vote, each party receives two seats, whereas, in a five‑member district, Party A would received three seats and Party B would receive two, a result that seems to better accord with people’s democratic intuitions.

[24].      See Steven J. Mulroy, The Way Out: A Legal Standard for Imposing Alternative Electoral Systems as Voting Rights Remedies, 33 Harv. C.R.-C.L. L. Rev. 333, 340 (1998) (highlighting the positive ramifications of choice voting, which minimizes wasted and surplus votes and thereby makes it easier for minorities to elect a second or third candidate).

[25].      See generally Jim Fitzgerald, U.S. Judge Selects Cumulative Voting To Protect Hispanics’ Rights in NY Village’s Elections, Star Trib. (Minneapolis-St. Paul, Minn.) (Nov. 6, 2009),
=69404622 (discussing a federal judge’s imposition of cumulative voting in a New York village of approximately 50% Hispanics that have never elected a Hispanic trustee or mayor).

[26].      Note that proportional voting in multimember districts produces good racial and ethnic diversity among elected officials while multimember districts using first‑past-the-post rules submerge minorities, dilute their voting strength, and undermine diversity. See, e.g., Thornburg v. Gingles, 478 U.S. 30 (1986); Tory Mast, History of Single Member Districts for Congress: Seeking Fair Representation Before PR, FairVote, (last visited Apr. 17, 2012) (describing how multimember districts and proportional representation would reduce problems inherent in single-member districts that consistently deny permanent minorities the chance for representation and would lead to positive results such as minority empowerment).

[27].      For a list of current U.S. congressmen from Maryland, see Maryland & the Federal Government, Md. St. Archives (Jan. 13, 2012),

[28].      See James D. King, Single-Member Districts and the Representation of Women in American State Legislatures: The Effects of Electoral System Change, 2 St. Pol. & Pol’y Q. 161 (2002); Richard E. Matland & Deborah Dwight Brown, District Magnitude’s Effect on Female Representation in U.S. State Legislatures, 17 Legis. Stud. Q. 469 (1992).

[29].      See Robert Richie & Steven Hill, The Case for Proportional Representation, Bos. Rev., Feb./Mar. 1998, available at
.html; Rob Salmond, Proportional Representation and Female Parliamentarians, 31 Legis. Stud. Q. 175 (2006).

[30].      For another argument in favor of single-member-district systems, see Mark E. Rush, Does Redistricting Make a Difference? (2000) (arguing that redistricting has little impact on voting results).

[31].       2 U.S.C. § 2(c) (2006).

[32].      See Mulroy, supra note 24, at 339.