The Case for Proportional Representation

by Robert Richie and Steven Hill

Nearly all elections in the United States are based on the winner-take-all principle: voters for the candidate who gets the most votes win representation; voters for the other candidates win nothing. This system is unjust and unnecessary. It is unjust because it leaves minorities (whether racial or political) unrepresented. As John Stuart Mill said, "It is an essential part of democracy that minorities should be adequately represented. No real democracy, nothing but a false show of democracy, is possible without it." It is unnecessary because we have immediate opportunities, at local, state, and national levels, to join the vast majority of mature democracies that have already adopted systems of proportional representation.

Proportional representation (PR) is based on the principle that any group of like-minded voters should win legislative seats in proportion to its share of the popular vote. Whereas the winner-take-all principle awards 100 percent of the representation to a 50.1 percent majority, PR allows voters in a minority to win their fair share of representation.

How does this work? A typical winner-take-all system of divides voters into "one-seat districts," represented by one person. With PR, voters in a constituency instead have several representatives: ten one-seat districts might, for example, be combined into a single ten-seat district. A party or group of voters that wins 10 percent of the popular vote in this district, then, would win one of the ten seats; a party or slate of candidates with 30 percent of votes would win three seats, etc. Various mechanisms work to provide proportional representation, and the details of different systems matter. But the principle of full representation is fundamental. Acceptance of it changes the way one sees electoral politics.

What's the Problem?

Consider three current failures of our winner-take-all system of representation:

* Members of racial and ethnic minorities are underrepresented;

* Voters' choices are restricted to candidates within the two-party, Republican/Democratic monopoly;

* Most legislative elections are effectively "no-choice" contests in districts dominated by a single party.

By restricting voters' choices and underrepresenting voters from minority groups, winner-take-all elections devalue the right to vote, our fundamental democratic right. Correcting these failures requires PR. No other political reform currently on the table--public financing of elections, term limits, fusion, or universal voter registration--will suffice to correct these deficiencies in our democracy.

Representation of Racial Minorities. At every level of government, the proportion of black, Latino, and Asian-American elected officials lags far behind these groups' share of the electorate. When members of a racial or ethnic group make up a majority of the electorate in a winner-take-all election, they tend to elect a member of their racial or ethnic group. Every majority-black US House district has a black representative; and in the 49 white-majority states, 144 of 147 US senators and governors are white. Most racial minorities clearly prefer representatives of their race, but winner-take-all elections often deny them a realistic opportunity to elect candidates of their choice. A quarter of our population is black or Latino, but these groups are in the minority in every state and as a consequence hold only one of 100 US Senate seats. The fact of such underrepresentation throughout our legislatures undercuts their legitimacy and effectiveness in addressing issues of concern to racial and ethnic minorities.

Two Parties. Winner-take-all elections prop up our two-party monopoly. Since 1960, new parties have formed at comparable rates in the United States and in European democracies using PR. But new parties in the United States are almost completely shut out of representation, whereas half the new parties in the European systems eventually have won seats--and the influence and organizing ability that comes from electoral viability. Polls show most Americans would like to see a third party electing candidates at every level of government, but only three of our nearly eight thousand state and congressional legislators were elected on a minor party ticket--all of them in Burlington, Vermont.

Minor parties by definition begin with minority support, which wins nothing in winner-take-all elections unless it is geographically concentrated. With little chance to win, minority party candidates cannot build or sustain support. Ross Perot's well-financed independent candidacy in 1992 won 19 percent of the vote, but he did not finish first in any congressional district. In 1996, his vote was reduced by more than half, although one voter in ten still voted for minor-party presidential candidates, and half of all eligible voters saw no reason to participate.

No-Choice Elections. One-party dominance of most American legislative districts provides a more subtle, but more sweeping, indictment of our winner-take-all system. Most Americans, most of the time, experience "no choice" elections for city council, state legislature, and the US House of Representatives. In the last ten House elections, for example, more than 90 percent of incumbents were reelected. The average margin of victory in House races is consistently over 30 percent. More than one-third of state legislative races in the 1990s were not even nominally contested by both major parties; fully 68 percent were not contested in Massachusetts in 1996. So-called "swing" legislative districts feature genuine competition and a chance for voters to cast a meaningful vote, but they are exceptions.

The dominance of one-party districts should be no surprise: gerrymandering allows legislators to choose their constituents in redistricting before their constituents go to the polls to choose representatives. Even though political intentions can be removed from, the redistricting process--as in Iowa's criteria-driven procedure, for example--its political effects are unavoidable. Given that some (perhaps most) districts will be non-competitive in winner-take-all elections, all districting ends up being a form of gerrymandering.

The ramifications of our fundamentally lopsided political landscape are often ignored in debates over term limit proposals and campaign finance reform. The real culprit for non-competitive elections is winner-take-all elections, not incumbency and inequities in campaign spending. In most districts, a clear majority of voters prefers one party's political philosophy to that of the other party. Consider open-seat elections, with no incumbent competing for the seat, and none of the financial advantages that come with incumbency. In 1996, Republicans won 29 of 35 open House seats in districts where Bill Clinton ran behind his national average, despite being outspent in a third of their victories. Yet Republicans won none of the 18 districts where Clinton ran ahead of his national average, despite being financially competitive in half of those defeats. This trend is not confined to elections in presidential years. Overall, Democrats hold 99 of the 100 US House districts where Clinton ran most strongly in 1996. Of the 150 districts where he ran most weakly, Republicans hold 134.

To be sure, congressional winners usually outspend their opponents. But that is because money follows power: to gain access, most major campaign contributors invest in candidates they expect to win. The great majority of voters are consistent in their voting patterns both between and within elections. We should be relieved that voters are well-grounded in a political philosophy, but frustrated that this consistency leads to most of them experiencing no-choice elections.

Why PR?

Support for PR as an alternative to winner-take-all politics has come from a diverse and distinguished group, including Alexis de Tocqueville, Charles Beard, Walter Lippman, Jane Addams, A. Philip Randolph, Robert Kennedy and, quietly, Franklin Roosevelt. The most outspoken early supporter of PR was John Stuart Mill, in his Representative Government (1861)--written less than two decades after the first works detailing possible PR systems.

The Majoritarian Argument

Perhaps Mill's most important contribution to the case for PR was his argument that majority rule itself is improved by full minority representation. By maximizing the number of voters who elect candidates, he pointed out, PR increases the chances that a legislative majority has support from a majority of voters; it is required for full representation, with voters having the power to elect representatives reflecting a range of opinion; and it fosters a deliberative legislative process which improves the majority view by ensuring that minority opinions are represented and heard.

As Mill observed, any particular majority is a collection of minorities, not a monolithic bloc. Once some voters are excluded from representation, policy can be passed without the support of a majority of the electorate. Suppose, for example, that all representatives win their elections with only 50.1 percent of votes. A law passed with support from only 50.1 percent of the legislators then would have backing from only a quarter of votes cast. Mill's point is no mere theoretical concern. In the 1994 "Republican revolution," in which Democrats lost their 40-year stranglehold on the US House of Representatives, fewer than one in four eligible voters voted for a winning House candidate. As a result, House passage of any particular bill in 1995 required the votes of representatives elected by only 13 percent of eligible voters.

By contrast, legislation in democracies with PR generally requires the support of representatives elected by a far higher percentage of the electorate. In Germany's 1994 elections with PR--with a high turnout and a high percentage of effective votes typical of European PR elections--more than 3 in 4 eligible voters elected candidates. (4 in 5 eligible Germans participated, and 19 in 20 voters elected a representative.) So passage of a bill required the votes of representatives elected by nearly 40 percent of eligible voters.

Majority rule also is undercut by winner-take-all elections because they drive voters into two camps. But two-choice elections obscure shades of difference and create the illusion of majority support for the winner. Mill stressed the importance of voters having a full range of choices and representation of their different communities of interest. "I cannot see," he wrote, "why the feelings and interests which arrange mankind according to localities, should be the only ones thought worthy of being represented." The notion that geography should be the primary basis of representation is even more antiquated now, given the increased mobility of our population, ease of communication across distances, and importance of economic, social, and political associations without geographic definition.

Finally, PR is important for majority interests because, as Mill argued, it provides represented minorities with a platform to challenge conventional wisdom. An advocate of universal suffrage, Mill still was sympathetic to conservative concerns about educated minorities being outvoted by newly enfranchised, less-educated voters. Assuring a voice to the minority eliminated his fears because of his faith in the results of a fully democratic process, with open and organized discussion among competing political ideas and projects. By allowing dissenters to win representation, PR fosters ongoing challenges to majority opinion, and thus complements our First Amendment freedoms.

In conjunction with attack ads, polling, and focus groups, the system of winner-take-all elections has made it extremely difficult to have reasoned political debate on certain contentious issues. These issues can assume great symbolic weight for swing voters--ironically, because they are among the relatively few voters with so little political grounding that they will support either party. The death penalty, for example, has come to represent "toughness" on crime. Because winner-take-all elections make nuanced positions difficult, and require that candidates win the support of politically indifferent swing voters, opponents of the death penalty find it hard to run credible campaigns for president or for most legislative offices. As with a whole range of issues--from drug policy to abortion rights to welfare reform--debate in political campaigns tends to lock into place, making it that much more difficult to challenge public opinion.

Mill's majoritarian argument for PR gains empirical support from a recent statistical comparison of 12 democracies in Europe. John Huber and G. Bingham Powell contrast a "Proportionate Influence Vision" of democracy, in which "elections are designed to produce legislatures that reflect the preferences of all citizens," with the "Majority Control Vision," in which "democratic elections are designed to create strong, single-party majority governments that are essentially unconstrained by other parties in the policy-making process." They conclude that "governments in the Proportionate Influence systems are on average significantly closer to their median voter than are governments in the Majority Control and Mixed systems. . . . If voters are presented with a wide range of choices and electoral outcomes are proportional, governments tend to be closer to the median."

In short, governance is more likely to take place at the center of the political spectrum with PR, since the electorate is fully represented and voters are able to express a wider range of preferences. At the same time, fair representation of the margins provides a mechanism to transform policy by shifting the political center. Opposition voices will be heard, and their ideas will be far more likely to be debated. If those ideas win growing support, the major parties will adjust accordingly in order to hold onto their supporters.

Other Reasons for PR

Mill's majoritarian argument is not the only case for PR. Four other claims are commonly offered in its support:

1. PR increases voter turnout. Voter turnout is generally estimated to be 10-12 percent higher in nations with PR than in similar nations using winner-take-all elections. This difference is understandable. In the United States, as we indicated, relatively few legislative elections are competitive, and our analysis of recent House elections demonstrates a strong correlation between the degree of competition and the level of participation.6 People in non-competitive districts--whether supportive of the majority or minority--might better invest their time and resources by supporting candidates in competitive races elsewhere than by voting in their own.

In PR systems, winning fair representation is dependent on voter turnout. Because nearly every vote will help a party win more seats, voters have more incentive to participate, and parties have incentives to mobilize their supporters. Moreover, parties and other electoral organizations have strong incentives to keep their supporters informed, and informed citizens are more likely to vote.

2. PR provides better representation for racial minorities. The 1982 amendments to the Voting Rights Act resulted in more districts drawn with majorities of racial and ethnic minorities. This increase in "majority-minority districts" produced a remarkable leap in representation of people of color in the US House in 1992. Between 1990 and 1992, the number of black and Latino House members jumped from 35 to 55.

In a series of recent rulings against so-called "racial gerrymandering," the Supreme Court has made it much harder to establish majority-minority districts; the result is almost certain to be a decrease in the number of elected black legislators. Lani Guinier, Jesse Jackson, and other civil rights leaders have argued for PR as an alternative, and already more than 75 localities have adopted semi-proportional systems to settle voting rights cases.7 By building from a fundamental principle of political fairness, PR could secure voting rights of racial minorities, without specifically targeting minority voters (just as Social Security protects low-income seniors by providing benefits for all).

In addition to winning a fair share of seats, minorities would have greater opportunities to negotiate for influence because they could "swing" among parties. South Africa used PR in its first all-race elections in 1994, and the two leading parties--the African National Congress and the National Party--ran multiracial slates with messages of inclusion. When New Zealand had its first PR election in 1996, the first Asian citizen was elected, and Pacific Islanders and indigenous Maoris tripled their representation. A Maori-backed party formed a coalition government with the governing party--a party whose relationship with Maoris had been analogous to Republicans' post-1960 relationship with American blacks.

By improving representation, PR in turn encourages minority communities to mobilize and win access to power. From 1925 to 1955, Cincinnati used the "choice voting" form of PR to elect a nine-seat city council. (See sidebar for an explanation of choice voting.) In 1929, when blacks were barely 10 percent of the population, a black independent candidate ran a strong campaign. In the next election, he was added to the Republican party's slate and was elected. In 1947, when blacks were 15 percent of the population, a former president of the Cincinnati NAACP ran in large part to defend the choice voting system that was under attack from Republicans seeking to restore their old domination of the council. In an indication that any substantial group of voters cannot be ignored with PR, the other major party slate supported him in 1949. He was elected, resulting in black representatives holding two of nine seats.

3. PR increases the number of women in office. The percentage of women elected to office in the United States--only 11 percent of the US Congress--is scandalously low, particularly in light of the relative strength of the American women's movement compared to other nations with far higher percentages of women legislators. Studies show that women representatives make a qualitative and quantitative difference in the type of legislation introduced and passed, yet the growth of women in state legislatures and Congress has stalled since 1992 despite relatively high turnover and the historic high in women's candidacies in 1996.

In state legislative elections, women win seats in significantly higher percentages in multiseat districts than in one-seat districts. The major reasons for this difference are that women are more likely to run and voters are more likely to seek gender balance when there is more than one seat to fill. Because PR expands options, PR systems give women additional leverage to force the major parties to support more women candidates. In 1994, a threat by women supporters of the major parties in Sweden to form a new women's party led to women winning 41 percent of seats in 1994 because the major parties recruited more women candidates. New Zealand, Italy, and Germany are among a growing number of democracies that use systems with a mix of winner-take-all districts and PR seats. It is instructive that women in all three countries are three times more likely to win seats elected by PR than to win in one-seat districts.

4. PR ends gerrymandering. Drawing district lines for political purposes has occurred from the first redistricting--the term "gerrymander" refers to a Massachusetts district plan drawn in 1812. But gerrymandering has become far more potent in an era of powerful computers, more detailed census information, and better techniques for measuring voter preferences.

As one example, Democrats in control of the redistricting process in Texas in 1991 placed the eight Republican incumbents in districts that were packed to be among the most conservative in the nation. These incumbents were easily reelected in 1992, but Democrats won 21 of the remaining 22 seats with only 50 percent of the statewide vote. Only one race was won by less than 10 percent, and the three open seats went to state legislators serving on redistricting committees. Congresswoman Eddie Bernice Johnson, the primary architect of the plan, admitted in 1997 that the redistricting process "is not one of kindness. It is not one of sharing. It is a power grab."

PR makes gerrymandering of any sort far more difficult. The smaller the percentage of votes that can be "wasted" on losing candidates--49 percent in a winner-take-all race, but less than 20 percent in a five-seat PR election and less than 10 percent in a 10-seat PR election--the harder it is for legislators to manipulate electoral outcomes.

Progressives Need Multiparty Politics

The case for PR is fundamentally non-partisan. Voters across the spectrum can support greater democracy or feel poorly represented by winner-take-all elections. But American political progressives have a particularly urgent need to support PR because of the growing problems created by the lack of a serious electoral vehicle to the Democrats' left. Many progressives overrate their current degree of support in the electorate, while others leap in equal error to desperate conclusions about the electorate's likely conservatism. The more complicated reality is disguised by winner-take-all elections, which divide voters into two camps and leave much progressive thought on the margins of political dialogue and influence.

A progressive party that won 5-10 percent of legislative seats in PR elections could have a great impact on public discourse, cross-issue organizing, and the policies of the Democratic Party. Electorally, this party could check the Democrats' rightward drift. It would give the base of the Democratic Party--labor, African Americans, feminists, environmentalists, defenders of civil liberties--a credible alternative, and thus an influence within the Democratic Party similar to that now held by the relatively few "swing" Democrats willing to vote for Republicans. When all voters, not just the centrist fringe, can swing their votes, the major parties must pay more attention to their base.

In addition, a progressive party would provide an opportunity for progressive activists to join together to build an infrastructure of independent politics--work that is very difficult to pursue in a two-party system that pushes activists into single-issue politics. It would give progressives greater access to media and an ongoing means to challenge conventional wisdom. It is one thing for the media to ignore activists who have no strong supporters in Congress; it would be more difficult to ignore a congressional contingent that consistently demonstrated a credible level of national support.

A progressive party would force healthy, if difficult, policy debates among progressives. Winning 10 percent of the vote in a PR system is hard work. The German Greens have never reached 10 percent nationally, yet have had a remarkable impact on German policy and the environmental positions of both major parties. They are partners in governing coalitions in several German states and may be part of the next national government. At the same time, Greens have made great strides in building coalitions among their members. Organizing a party need not replace other grassroots organizing: a study of German Greens found that over 80 percent of members were active with an organization outside the Greens. But having the Green Party as a unifying electoral presence has made their other work more effective. American progressives urgently need PR to create a similar unifying electoral vehicle.

With PR, a progressive party also could find unexpected allies. Progressives might not outdo conservatives in winning over the majority. But a winner-take-all, two-party system facilitates "divide and conquer" strategies in which a conservative party can cut into the potential economically progressive majority with such wedge social issues as gun control, gay rights, affirmative action, and abortion.

In a two-party system, conservatives can create an electoral majority with a set of positions that can be opposed by the majority of voters, but bring together fervent minorities willing to accept positions they oppose in exchange for support for their issue: social liberals seeking low taxes, blue-collar Catholics opposing abortion, labor union members opposing gun control, and so on. With electable choices across the spectrum, a multiparty system based on PR would allow us to find out where the American people really stand--and on many issues, they arguably will stand to the left of current policy. The political center of most of Europe, with its policies on health care, welfare, worker rights, and the environment, is where American progressives would love to be.

Why Now?

Many reformers will quickly accept theoretical arguments for proportional representation, but question the viability of a PR movement. Some mistakenly think PR would require constitutional change or demand overly dramatic changes in our political culture. Others confuse PR with parliamentary government, although PR directly affects only how one elects a legislature, not governmental structure. Forms of PR could work extremely well with simple statutory changes, and a confluence of events in the 1990s provides a remarkable opportunity to work for PR's adoption. These developments include:

1. Winner-take-all politics cannot be fixed. We have a particularly realistic opportunity to promote PR because of how well it addresses widely accepted failures in winner-take-all politics. Some of the most egregious problems reflect irreversible changes in technology, campaign techniques, and demographics. Without PR, no political reform--including the best of campaign finance reforms--can prevent most campaigns from being developed from focus groups of swing voters rather than principled policy positions. Campaign consultants know too much about how to win elections under winner-take-all rules. By freeing the majority to elect candidates they want, PR would weaken the stranglehold of the swing voters that give campaign consultants such power in winner-take-all elections.

2. Other reforms face barriers. The most electorally successful political reform movement of the past decade has been the effort to put limits on the number of terms that legislators can serve. The Supreme Court has quashed term limits for congressional elections and may follow suit for state legislative elections. But the goals of term limit supporters are in any case only partly achieved by limits: most voters continue to live in one-party districts and to be frustrated by poll-driven politicians. Moreover, for all the voter disgust with money in politics, campaign finance reformers now seem to be at an impasse at the federal level. The Buckley v. Valeo ruling, independent expenditures, and wealthy self-financed candidates bedevil reformers. Public financing supporters are making an expensive, potentially historic effort at the state level, and soon we will see how voters respond. But some analysts are skeptical that public financing can win in many states in the current climate of political depression. And even if it succeeds, the fundamental injustice of winner-take-all elections will remain. As Lani Guinier points out, it is not enough to take the money out of elections; we need PR to put the people in.

In any event, PR is an attractive complement to other reforms. And all of the reform energy that has developed in recent years provides an infrastructure of support for PR campaigns.

3. The opportunity to build a powerful coalition. As already discussed, women, racial minorities, advocates of term limits and campaign finance reform, minor party supporters, and progressive constituencies within the Democratic Party all have particular incentives to support PR. Republicans also are facing growing splits, particularly on social issues, and losers in those internal Republican debates may be ready for opportunities to maintain and build representation with PR. The Center for Voting and Democracy is developing working relationships with representatives of most of these constituencies, who together form a majority coalition similar to those that developed in recent successful campaigns for PR in New Zealand and Scotland.

Getting Started

Efforts to bring PR to American elections build on a rich history. Earlier this century, two dozen cities, including Cincinnati, Cleveland, and New York, adopted the choice voting method of PR by initiative. Today's movement can learn much from this early PR movement. Choice voting was successful in achieving its reformers' primary goal: undercutting the power of one-party political machines. Unfortunately, this success led to these machines' unrelenting hostility. Although only two of the first 26 attempts to repeal choice voting in cities around the nation were successful, the previously dominant political forces eventually outlasted reformers and won repeals everywhere except Cambridge, Massachusetts. But the primary vehicles of anti-PR attacks--racist and anti-communist appeals and concerns about costly electoral administration--can be addressed far better today.

PR activism is on the rise again. Rep. Cynthia McKinney has introduced the Voters' Choice Act (HR 3068) to restore the option states had before 1967 to elect their Congressional delegations by PR. The bill is acquiring a growing number of co-sponsors, and other pro-PR legislation likely will be introduced in Congress in 1998. Several state groups have formed to promote PR, and recent PR initiatives in two major cities--Cincinnati and San Francisco--won 45 and 44 percent of the vote, respectively, despite limited funding and media exposure.

To be sure, many Americans--particularly elected officials--may be cautious about moving away from our political traditions. Moreover, much voter attention will continue to focus on "single-winner" executive offices--such as president, governor, and mayor--which do not allow for PR, since PR requires multiseat districts. Even so, there still are immediate opportunities to reform plurality methods that are used in most states.

Instant Runoffs

For executive offices, Australian-style instant runoff voting (IRV) would provide both better majority representation and minority participation than plurality voting. Australia uses IRV for parliamentary elections, Ireland uses it to elect its president, and the United Kingdom may well adopt it within two years in a national referendum on parliamentary elections.

With IRV, voters rank candidates in order of choice: 1, 2, 3, and so on. Each voter still has only one vote, but ranking candidates allows the ballot-count to simulate a series of run-off elections. If no candidate wins a majority of first-choice votes, the last-place candidate is eliminated. Ballots cast for that candidate are redistributed to each voter's next choice. This process of elimination occurs until a candidate wins majority support.

In the many states and localities still using traditional runoffs for primary or general elections, IRV would save money for taxpayers and campaign cash for candidates by combining two elections into one. Moreover, elected officials can appreciate IRV because it eliminates the "spoiler" problem created by minor parties--a problem for the major parties that the Reform Party and Green Party show great interest in expanding. And for minor parties, IRV reverses the "wasted vote" calculation. IRV allows minor party candidates to participate fully and potentially build their party's support. At the same time, Australia's experience with IRV demonstrates that it gives a minor party some leverage over major parties. A minor party candidate can call on supporters to hold back from casting second-choices for a major party candidate unless that candidate agrees to support some of the minor party's issues.

States that likely will debate IRV legislation in 1998 include Alaska, New Mexico, Pennsylvania, Texas, and Vermont. An IRV state initiative may be launched in California in 1998; IRV could even be adopted for presidential elections by state legislation or initiative.

Three-Seat Districts

City, state, and national efforts for instant runoff voting would complement another modest modification of our rules that has the potential to draw support from the current political establishment: three-seat legislative districts with a 25 percent victory threshold.

From 1870 to 1980, Illinois used the semi-proportional system of "cumulative voting" in three-seat districts to elect its lower house. Illinois elections were a modest departure from winner-take-all elections. Voters had three votes, but had the option to put all three votes on one candidate. If 25 percent of voters supported only one candidate, that candidate was sure to win. That is the mathematics of PR in a three-seat district: just over 25 percent wins one seat, just over 50 percent wins two and over 75 percent of votes is necessary to sweep the district.

This relatively minor modification of winner-take-all rules had a profound impact on Illinois politics. Perhaps most significantly, nearly every constituency had two-party representation. Although most one-seat districts now are safe for one party, both in Illinois and around the nation, there are relatively few areas where at least 25 percent of voters are not ready to support another party--Bill Clinton won at least 25 percent of the vote in all U.S. House districts in 1996. A semi-proportional system gives these minority voters a chance to win representation.

In Illinois, most constituencies typically had two representatives reflecting two major factions within the majority party and one representative from the smaller party. These minority-backed legislators played a creative role in the legislature. In 1995 the Chicago Tribune editorialized in support of cumulative voting's return, writing that "[M]any partisans and political independents have looked back wistfully at the era of cumulative voting. They acknowledge that it produced some of the best and brightest in Illinois politics."

The Center for Voting and Democracy recently commissioned a study of Illinois' use of a semi-proportional system. Interviews with Illinois political leaders show strong bipartisan support for cumulative voting, including support from the state senate's majority leader and minority leader. A recurring theme is that semi-PR systems in three-seat districts actually provided better geographic representation than smaller, one-seat districts with monopoly representation. Constituents had more options when provided with access both to representatives in the majority party and the minority party. And both parties had direct interests in serving the needs of all parts of the state. The loss of cumulative voting has meant loss of bipartisan support for policies of particular interest to one-party strongholds; Chicago, for example, has been a big loser in equitable funding of public schools.

Three-seat districts are particularly promising as an alternative means to enforce the Voting Rights Act. If North Carolina adopted a proportional system with four, three-seat districts for U.S. House elections, it would likely result in a greater number of competitive black candidates than the controversial redistricting plan that triggered several rulings against race-conscious districts around the nation. Such plans in the south also would increase representation of women and white moderates. Of 36 House members in the deep south (South Carolina, Georgia, Alabama, Mississippi and Louisiana), only one is a woman and only four are white Democrats.

Though cumulative voting has attractions, it also suffers from serious flaws, so we prefer either choice voting or an open party list system in such three-seat plans. Choice voting allows voters to rank candidates and have their votes coalesce behind candidates with the strongest support within major constituencies. In an open party list system, all votes for candidates from a particular party's "team" of candidates will boost the party's chances of winning seats.

As a first step toward change, we urge that states form commissions on reapportionment. These commissions could address issues such as: the number of legislators in a state; the problems deriving from the increased ability to gerrymander lines; and the potential of semi-PR and PR plans. An influential commission established in the mid-1980s in New Zealand surprised most political leaders by recommending replacement of the nation's winner-take-all system with a fully proportional system; that system was adopted by the voters in 1993 despite intense and well-financed opposition. Given the power of the argument for at least some modifications of winner-take-all elections, any commission with a degree of independence may well generate surprises in the United States as well.

Some Objections

To conclude, we consider some familiar objections to PR systems.

1. Political instability. True, Israel and Italy both use forms of PR and have faced problems of governmental instability. We could explore the historical and political complexities of these countries and their systems in order to exonerate the principle of PR, but suffice it to say that most mature democracies with PR are not plagued by falling coalitions or right-wing religious parties.If we don't condemn winner-take-all elections by citing Algeria, Pakistan, and India, then why condemn PR by citing Italy and Israel? There are currently 36 nations with more than two million people and high 1995 ratings from the human rights organization Freedom House. Of those 36, fully 30 use PR to elect their most powerful legislature, while only three--the United States, Canada and Jamaica--elect all national bodies with a winner-take-all system.

2. Excessive gridlock. Some argue that we have enough gridlock with two parties and that adding more to the mix will simply make things worse. One answer to this concern again is empirical: nearly every major democracy has more than two-party representation, and most are not paralyzed by gridlock. In fact, many PR democracies--including Germany, Sweden, Netherlands, and Switzerland--have developed far more comprehensive policy than the United States on such major issues as health care and immigration. A two party democracy rewards constant mud-slinging and obstructionist posturing because if one party can drive up the negatives of the other, voters have only one place to go. "Zero sum politics" translates into "zero sum governance."

3. Loss of district representation. The advantage of district representation, it is said, is that all areas have someone to hold accountable for district issues. The problem is that most residents don't vote for their representatives and can't even identify them.

Furthermore, voters can take little comfort from being represented by someone who is sharply opposed to their own political philosophy. PR takes a different approach. All voters deserve an opportunity to choose a representative who thinks like them. With PR, voters find an ideological "home" rather than a geographic one. Their choice of representation may be influenced by local considerations, and systems can be designed to ensure some geographic representation, but geographic interests are not assumed to be paramount.

These three objections to PR do not exhaust the conventional list of concerns. But virtually every additional objection to PR, like those addressed here, is founded on the insulting theory that voters cannot handle the demands of making real choices. The typical winner-take-all advocate wants to keep things simple for the "poor bastards," who, left to their own devices, will keep electing unworkable governments and dangerous extremists. Empirical study and democratic principle condemn this charge. It is as objectionable as arguments against full suffrage.

We are skeptical, too, about arguments based on American exceptionalism. Yes, we are a continental democracy, with a unique constitution that makes accountability difficult, and long-standing traditions that should be modified with care. So we need to think hard about how best to realize the moral and political imperative of full representation. But that imperative itself retains its full and compelling force.

Committed democrats should act on it, and the ideal opportunity is quickly emerging. The redistricting process is the Achilles heel of our winner-take-all system. Behind closed doors, once every decade, the duopoly carves up the electorate, leaving most of us with another decade of no-choice legislative elections. The next gerrymandering is set for 2001. With 50 states as potential battlegrounds and voter frustration everywhere, a movement for PR has a perfect opening and a natural rallying cry that fits with its own democratic impulse: "This time let the voters decide."

Politics watch