Full Representation: Uniting backers of gerrymandering reform and minority voting rights

By Robert Richie
Published November 1st 2004 in National Civic Review
In April 2004, the Supreme Court issued a decision with little fanfare that may have more impact on voter choice and representation in the U.S. Congress than any single event during the past decade. In Vieth v. Jubilirer1, the Court upheld Pennsylvania’s congressional redistricting plan despite all justices agreeing it was a partisan gerrymander. In a divided court – there were five separate opinions – the majority saw no constitutional violation, and the inherent partisan consequences of any judicially mandated reform suggest that future courts will also have difficulty rejecting political gerrymanders.  The decision means changes in the status quo now will only come through the political process: that’s you, me and our representatives elected from gerrymandered districts.  

Yet, at the same time that the high court is accepting partisan gerrymanders, it has undercut efforts to diversity legislatures through redistricting. Although not as devastating for representation of people of color as initially feared, the line of cases starting with Shaw v. Reno2 in 1993 have made it harder for racial minorities to increase or even maintain political representation – and doing so would be significantly harder if states prohibited use of political data in districting. These legal setbacks for pro-democracy goals, however, provide a unique opportunity to unite opponents of gerrymandering with advocates for minority voting rights in support of a win-win solution: full representation voting methods in non-winner-take-all, multi-seat legislative districts.

Designer Districts

Does redistricting matter? You bet it does. Take Virginia, where Democrats in 2001 won their first statewide race for governor since 1989. But Republicans went from barely controlling the statehouse to a two-thirds majority for one simple reason: for the first time since Reconstruction, Republicans had been able to draw the district lines before the election. In Florida, Democrats are strong enough to hold both U.S. Senate seats and gain a virtual tie in the presidential race. But with full control of rigging the district lines, Republicans hold an overwhelming 18 of 25 seats in the U.S. House of Representatives. In 2003, Republicans infamously opened up the potential to gerrymander congressional districts mid-decade. If the Supreme Court accepts the Texas plan – and its potential to help Republicans gain seven seats -- we could see incumbent legislators lining up to fine-tune their districts after every election.

The 2001 redistricting already was perhaps the most flagrantly rigged insider's racket in decades. In California, for example, incumbent U.S. House Democrats paid $20,000 apiece to a redistricting consultant to have "designer districts" drawn for them. Republicans went along with this cozy arrangement in exchange for their own safe seats. The result was an unbroken parade of landslide wins in 2002, with no challenger to any of the 50 incumbents winning even 40 percent of the vote. Nationally, only four challengers defeated House incumbents, the fewest in history, while fewer than one in 10 races were won by competitive margins inside 55 percent to 45 percent. Perhaps most perniciously, it’s a simple exercise to predict who will win easily; two days after the November 2002 elections, our Center’s biennial Monopoly Politics report3 projected winners in more than 80 percent of the House races in November of 2004 based on a model that has been accurate in all but one of more than 1,250 projections since 1994.
Unyielding Control and Distorted Representation
The lock-up of the U.S. House has major repercussions for our political process and for representative government. Elected every two years, with representatives closer to the people than senators or the president, the House was designed to reflect the will and different interests of the nation. The reality is far different. The growth in seats held by women has come to a standstill, stuck at 14 percent. Hardly any members can be held accountable at the polls, given the lopsided general elections and the paucity of meaningful primary challenges. The result is a highly partisan Congress, with most representatives’ voting records closely correlating with the desires of their party leadership.
Control of the House is nearly as fixed in stone. Since 1954, control of the House has changed just once, when Newt Gingrich and Republicans took over after the 1994 elections. Democrats gained a few seats in each election between 1996 and 2000, but Republicans cemented their grip in 2002 after dominating redistricting in several large states. In 2000 Democrat Al Gore won more combined popular votes than Republican George Bush in the four states of Florida, Michigan, Ohio and Pennsylvania, but after having unfettered control of redistricting in those states, Republicans now hold a whopping 51 out of their 77 seats.  Despite Gore’s national edge in the popular vote in 2000, Bush carried 241 of current House districts compared to only 194 for Gore. Redistricting contributed to this bias, but a bigger reason is where the parties gain their support, with the Democratic vote much more concentrated in cities. As a result, despite Democrats theoretically needing to pick up only 12 seats to regain the House, few observers believe that will be possible this decade without a dramatic national voter surge. Furthermore, a win for Bush in 2004 would likely lead to a wave of retirements of Democratic House members, whose only chance at influence is a sympathetic president, and make it even harder for Democrats to win a majority.

State legislative elections are even less competitive than U.S. House races. Of the thousands of state legislative races over the last three election cycles, a whopping 40 percent were uncontested by one of the two major parties because the districts were so lopsided it was considered a waste of campaign resources for the minority party to contest for these seats. That's two in five races where the only choice for voters was either to ratify the candidate of the dominant party or not vote at all. As to partisan control, not a single legislative chamber changed hands in 2002 in a state where incumbents controlled the redistricting process -- even as half of the 36 gubernatorial races in the same election resulted in a partisan change.

Districting also has a huge impact on representation of racial minorities. Increases in representation of people of color in our state legislatures and U.S. Congress have slowed dramatically. The number of African Americans, Latinos and Asian Pacific Americans in the U.S House nearly doubled from 30 in 1983 to 59 in 1993 – largely due to a sharp increase in the number of congressional districts in which racial minorities comprised a majority of adults in a district.  But from 1993 to 2003, the period immediately following the Shaw v. Reno U.S. Supreme Court ruling, the number of racial minorities only increased from 59 to 61 – with an actual decrease in the number of African Americans and Asian Pacific Americans.  On the whole, racial minorities make up less than 15 percent of the U.S. House, despite representing almost a third and growing share of our nation’s population, yet for the foreseeable future, only Latinos appear likely to see a noticeable increase in electoral opportunities through districting. Given demographic shifts, African American representation may well decline in the coming decade, as it already has in California state legislative districts – particularly if the Voting Rights Act is weakened in 2007 when key provisions are set to expire State legislatures show similar trends.

Fair Shares
So, what now? Congress has the power to regulate congressional redistricting and require states to establish nonpartisan commissions that draw lines based on clear criteria. Not doing so is analogous to allowing elected officials to count votes in their own elections behind closed doors, boosting their friends and hurting their enemies. States also can take action. Indeed, groups like Common Cause are contemplating a wave of state ballot measures. At the very least, the redistricting process should be a very public one, with full disclosure, more media coverage and frequent opportunities for citizen input.

Redistricting reform alone, however, would have limited impact on electoral competition, which is perhaps the dominant concern among many political reformers, and often would not address the partisan bias that hurts whichever party – typically Democrats – has its base of support more concentrated in cities. More importantly, it could easily clash with the goals of civil rights organizations seeking to increase representation of racial minorities, as their strategies often depend on using political data as a substitute for racial data and on drafting lines that stand out.
What we ultimately must confront is our exclusive reliance on winner-take-all elections. Winner-take-all elections, whether in single-member districts or for at-large positions, require winning candidates to attract a majority or substantial plurality of the vote. By definition, candidates representing political minorities have great difficulty amassing such a large a share of votes and therefore stand little chance of being elected. The same holds true for minority candidates running in racially polarized districts, as evidenced by the fact that the 49 states where white voters are the largest group have 97 white U.S. Senators and the one state (Hawaii) where white voters are not the largest group has two Asian American Senators.

Under our current system, racial minorities and the poor have the right to vote but are often denied the equally fundamental right to representation. Most enduring democracies have rejected the winner-take-all model in favor of systems that ensure that any grouping of like-minded people—minorities and majorities—gets a fair share of power and representation in legislative bodies, whereas our current winner-take-all principle can award 100 percent of the representation to a 50.1 percent majority. If African American voters comprise 20 percent of the vote in a racially polarized county, fair voting systems would allow them to elect a representative to at least one of the five seats—rather than be shut out, as they would be in a traditional at-large election or in a single-member district plan that dispersed their vote across several districts.

A win-win for women, racial minorities and supporters of more partisan fairness and more competitive elections, full representation could be adopted for nearly all legislative bodies in the United States—including most state delegations in the House of Representatives—without having to amend the Constitution. One example consistent with American traditions comes from Illinois. For more than a century Illinois voters elected their state legislature with a full representation voting method called cumulative voting, with candidates running in bigger districts that each had three representatives. Lowering the victory threshold for candidates from 50 percent to 25 percent did not overturn the two-party system, but it broadened representation within the parties, promoted more bipartisan policy and elected more women and people of color. The Chicago Tribune in 1995 editorialized that "Many partisans and political independents acknowledge that [cumulative voting] produced some of the best and brightest in Illinois politics." 4

More recently, in May 2000, the citizens of Amarillo, Texas, filled four seats on its school board for the first time by cumulative voting. No black or Latino candidate had been elected to the board in more than two decades, despite Latinos and African-Americans making up more than 20 percent of the city’s population and an even larger share of the student population. Instituted to settle a voting rights lawsuit in 1999, cumulative voting had an immediate impact: a black candidate and Latino candidate won seats with strong support in their communities; voter turnout tripled over the most recent school board election; and all parties in the voting rights settlement expressed satisfaction with the new system. A person of color – one African American and one Latina -- was elected in the two subsequent cumulative voting elections in 2002 and 2004.

In the past 20 years, nearly 100 American jurisdictions have adopted a full representation method to settle voting rights challenges, and federal judges several times have sought to impose them directly as remedies in voting rights cases. Perhaps the fairest of these systems, the choice voting method recommended as an option in the National Civic League’s Model City Charter, has been used for decades to elect the city council and school committee in Cambridge, Massachusetts. Cambridge is famous for feisty local elections, good government and higher voter turnout than its neighbors and for decades has provided fair representation for African Americans. Fair racial representation was also typically true when choice voting was used to elect city councils in New York City, Cincinnati and other major cities before their repeal in the Cold War climate of the post-World War II era.

Significant organizations have grown to support full representation voting methods. In 1998, a National Black Caucus of State Legislators task force found strong interest among black legislators. The League of United Latin American Citizens (LULAC), National Association for the Advancement of Colored People (NAACP) and the Mexican American Legal Defense and Educational Fund (MALDEF) joined with local plaintiffs to win the adoption of cumulative voting in Amarillo, the largest city now using such a system. The National Conference of Black Political Scientists endorsed full representation in 1999. National and state affiliates of US PIRG, Common Cause, Sierra Club, National Organization for Women and the League of Women Voters adopted positions in favor of full representation.

This rise of interest in full representation in the 1990s obviously did not occur in a vacuum. Voting Rights Act provisions on redistricting divided and preoccupied the Supreme Court more than any other issue in the 1990s. The Court heard arguments in cases involving voting rights and redistricting nearly ever year in the decade, often in bitterly contested 5-4 decisions that had the general impact of limiting states’ use of race in drawing legislative district lines. Full representation methods solve many of the legal problems that arise in both political and racial gerrymandering cases, as has been pointed out by both Voting Rights Act backers like Lani Guinier5 and opponents like Supreme Court Justice Clarence Thomas.6 By boosting representation of people of color without the need for race-conscious districting, full representation avoids the legal tightrope created by the combination of the Voting Rights Act, which protects minority voting strength, and Shaw, which weakens the ability to draw districts that would enable that protection.  

Apart from legal battles over Shaw and philosophical concerns, civil rights attorneys have discovered, in states like Texas, Alabama, and North Carolina, that full representation can simply be a good fit with local conditions. Perhaps a minority community is more geographically dispersed than necessary for a single-member district plan. Perhaps a jurisdiction may want to avoid redistricting every decade. Perhaps there is frustration that most voters in a minority community are still left out of a chance to elect a candidate of choice even with a district plan that provides for enhanced minority representation. Perhaps in a multi-racial community, a citywide full representation plan is the easiest way for different racial minorities to elect representation.

Full Representation v Fairer Gerrymandering
It may seem like a leap to consider full representation as the defining strategy for tackling political gerrymandering, but its history overseas (the great majority of major democracies use full representation systems) and in Illinois state legislative races would suggest that it could work well in major American elections. With that in mind, compare what even a modest full representation system – for instance, choice voting in Illinois-style, three-seat districts -- would do for our politics compared to fairer gerrymandering (as indeed all single-member districts are gerrymanders for large numbers of people consigned to irrelevance during the course of the use of a particular map):

* Polarization: One of the great fallacies of the current debate over political gerrymandering is that polarization in Congress flows from redistricting. The evidence much more powerfully suggests that winner-take-all is the blame. States aren't gerrymandered, but the famous red and blue maps of political leanings show how predictable most states have become in a close national race because of greater partisan cohesion among voters. Prominent pollster Stanley Greenberg in his recent book estimates the number of true swing voters has decreased to about 10 percent of the electorate.  Fairer gerrymandering will not create that many more districts that are within the 45 to 55 percent range of partisanship. All the rest will remain quite predictable until partisan polarization changes in this country-- meaning that the number of competitive House races would likely at best  rise from one in ten seats to one in six under nonpartisan districting. In contrast, full representation would almost certainly mean two-party competition and representation in all districts -- a huge difference.

* Representation of racial minorities: Politically neutral gerrymandering does next to nothing for candidates of color – and sometimes less than nothing, as it quite likely will lead to fewer districts that are majority-minority. Full representation would increase opportunities for candidates of color in far more parts of the country -- and adding the element of choice voting would make it far more likely that people of color would win even when below the 25 percent threshold of exclusion, for there would be more meaningful candidacies of color.
* Representation of women: Fairer gerrymandering would only help women to the extent that with somewhat higher turnover (e.g., 95 percent of incumbents might win rather than 99 percent), there would be more opportunities. But full representation would have a far greater impact. Women historically have been much more likely to run -- and thus win -- in multi-seat districts systems than single-member district systems in American state legislative races. The evidence is overwhelming from international studies that women do better in full representation systems than in single-member district systems.
* Recognition of limits of our duopoly and drawing youth support: More than a quarter of Americans now register as independents, and they are disproportionately young. Fairer gerrymandering does nothing for independent candidates and for third party candidates. It does allow more independent voters to tip elections, but not most of them. In contrast, full representation, particularly with choice voting, puts independent voters in play around the country and gives third parties a fighting chance to run strong and hold major parties accountable. Illinois' full representation plan did not lead to multi-party democracy, but did open up representation to more independent voices in the major parties – independent political clubs were common. Use of choice voting would ensure that parties would have to work hard to be representative in a way that no single-member district plan will. There's a reason that full representation won 92 percent of the student vote in New Zealand’s national referendum on changing winner-take-all elections in 1993, with young people providing the margin of victory when full representation won overall with 54 percent of the overall vote.
If properly framed and explained and with greater support and resources, full representation has an exciting potential to bring together political reformers and civil rights advocates to support the same cause. It also can bring in women and young people in a way that campaigns for fairer gerrymandering cannot. Full representation also is winnable. An initiative to adopt choice voting for Cincinnati’s city council elections in 1991 lost only 55 to 45 percent despite a mere $15,000 campaign budget, opposition from the major daily paper and media exaggeration of the the potential costs of adopting the system. Choice voting also lost by only 12 percent in San Francisco in 1996 a mere 22 months after the idea was first raised in the city; the campaign budget also was relatively small at $25,000. Successes this decade for instant runoff voting, the winner-take-all variation of choice voting, and past wins for choice voting at the polls in the first half of the 20th century – including a landslide win in a stand-alone measure in New York City in 1936 -- also provide encouragement.

We can and should draw fairer districts and we can and should keep working for fair representation of racial minorities in current voting systems. But we must also look to full representation to achieve a simple goal: providing means to allow fair and realistic opportunities for voters to elect individuals of their own choosing. While no cure-all, they are a necessary step toward creation of a more inclusive, responsive political system.

[Robert Richie is executive director of the Center for Voting and Democracy, www.fairvote.org.]


Full Representation Voting Systems

• One-Vote System
    In the one-vote system (also called “limited voting”), voters cast one vote in a multi-seat race. In variations of the system, they might cast more votes, but still fewer than the number of seats; the greater the difference between the number of seats and the number for which one can vote, the greater the opportunities for minority representation. Versions of limited voting are used in Philadelphia, Hartford and numerous other local jurisdictions. It has been adopted to resolve at least 30 voting rights cases in Alabama and North Carolina since 1987.
    Example: In a race to elect five candidates, voters each cast one vote. Winning candidates are determined by a simple plurality: the five candidates with the most votes.

• Cumulative Voting
    In cumulative voting, voters cast as many votes as there are seats to be elected. But unlike winner-take-all systems, voters are not restricted to giving only one vote to a candidate. Instead, they can cast multiple votes for one or more candidates. Cumulative voting was used to elect the Illinois state legislature from 1870 to 1980. In recent years it has been used to resolve voting rights cases for city council and county commission elections in Alabama, Illinois and New Mexico and for school board elections in Alabama, South Dakota and Texas.

Example: In a race to elect five candidates, voters can cast one vote for five candidates, five votes for one candidate or any combination in between. The five highest vote-getters win.

• Choice Voting

Also known as “single transferable vote” and “preference voting,” choice voting is the most common candidate-based full representation system used in other nations. Each voter has one vote, but can rank candidates in order of choice (1, 2, 3, 4, etc.). Candidates win by reaching a “victory threshold” roughly equal to the number of votes divided by the number of seats. If a candidate has too little first-choice support to win, votes for that candidate are transferred to those voters’ next choices. This transfer of votes facilitates coalition-building and allows a candidate to run without fear of being a “spoiler” splitting the vote.

Choice voting has been used for city council and school board elections in Cambridge, Massachusetts, since 1941. Ireland and Australia use choice voting for national elections. The city council in Cambridge (where blacks are 13 percent of the population) has had black representatives since the 1950s. Choice voting in other Americans cities, including for five elections to the New York city council from 1937 to 1945, also resulted in fair racial, ethnic and partisan representation.
    Example: In a race to elect five candidates, voters can rank in order of choice as many candidates as they wish. Candidates win by gaining the support of about one-fifth of the voters. A ballot counts towards the election of that voter’s top-ranked candidate who needs that vote to win.


Given that redistricting has often been overlooked in our political analysis, here is a quick tutorial: Every 10 years after the release of new census numbers at the start of each decade, all legislative districts at local, state, and federal levels must be redrawn to make sure that they are approximately equal in population, which now means about 640,000 residents per U.S. House district. Some cities and states have procedures to promote the public interest in this redistricting process, but most do little to prevent the creation of a hodgepodge of districts gerrymandered to protect incumbents and build partisan advantage. The redistricting plan typically is passed like any legislative bill, by a majority in both houses and is signed by the governor. Whichever political party controls the line-drawing process in a given state has the God-like power to guarantee continuing majority control and to decide who will win most races. The ruling majorities rely on techniques such as "packing" and "cracking. “Packing” involves putting as many of your opponent's voters into as few districts as possible, sacrificing those districts but making all the surrounding districts more favorable to your side. "Cracking" means scattering opponent's voters into different districts, dividing and conquering as you go. With increasingly sophisticated computer software, polling results and demographic data that have made this process of unnatural selection much more sophisticated, incumbent legislators quite literally choose the voters before the voters have a chance to choose them


1 Vieth et al. v. Jubelirer, President of the Pennsylvania Senate, et. al. no. 02-1580, 2004.
2 Shaw v. Reno 509 U.S. 630,  1993.
3 Available at http://www.fairvote.org/2004/index.html.
4 Editorial, Chicago Tribune May 30, 1994 p.
5 Guinier, L. Tyranny of the Majority. New York: Free Press, 1994.
6 Majority opinion, Holder v. Hall 512 U.S. 630 1993.
7 Greenberg, S. The Two Americas. New York: Thomas Dunne Books, 2004.