Boston Review
table of contents
new democracy forum
new fiction forum
rave reviews
writers’ guidelines
bookstore locator
literary links


Search this site or the web Powered by FreeFind

Site Web


Second Proms and Second Primaries:
The Limits of Majority Rule

Lani Guinier

Brother Rice High School held two senior proms last spring. It was not planned that way. The members of the prom committee at Brother Rice, a boy's Catholic high school in Chicago, expected just one prom when they hired a disc jockey, picked a rock band, and selected music for the prom by consulting student preferences. Each senior was asked to list his three favorite songs, with the understanding that the band would play the songs that appear most frequently on the lists.

Sounds attractively democratic. But Brother Rice is predominantly white, and last year's senior prom committee was all white. That's why they ended up with two proms. The black seniors at Brother Rice felt so shut out by the "democratic process" that they organized their own prom. As one black student put it: "For every vote we had, there were eight votes for what they wanted. . . . [W]ith us being in the minority we're always outvoted. It's as if we don't count." Some embittered white seniors saw things differently. They complained that the black students should have gone along with the majority: "The majority makes a decision. That's the way it works."

In a way, both groups were right: with majority rule and a racially organized majority, "we don't count" is the "way it works" for minorities. In a racially divided society majority rule is not a reliable instrument of democracy. That's a large claim, and one I don't base solely on the actions of the prom committee in one Chicago high school.

In a recent voting rights suit in Arkansas, I represented some black plaintiffs in a case that turned, in the end, less on legal technicalities than on the relationship between democracy and majority rule. The failure to challenge traditional assumptions about that relationship -- to show that majority rule is sometimes unfair -- sealed the defeat of the plaintiffs. With the Arkansas case as background, I will discuss the standard remedy that courts use when black plaintiffs win in voting rights cases -- to establish a majority black electoral district. This strategy of "race conscious districting" is, I believe, an inadequate method for representing minority interests.

But if a group is unfairly treated when it forms a racial minority within a single district, and, if we cannot combat the unfairness by setting up a new district in which the racial minority is now a majority, then what is to be done? The answer is that we need an alternative to majoritarianism: a "principle of proportionality" that transcends winner-take-all majority rule and better accommodates the values of self-government, fairness, deliberation, compromise, and consensus that lie at the heart of the democratic ideal.

The Case of the Majority Vote Run-off

Phillips County is a predominantly rural, economically depressed county in Arkansas. Majority black in population, it is majority white both in voting age population and in registered voters. According to the 1980 census, 53 percent of the 34,772 residents are black; but blacks constitute only 47 percent of the county voting age population. Despite this representation in the population, blacks in the county have never had much political power. In fact, since Reconstruction, no black has ever been nominated to any county-wide office in Phillips County.

According to Sam Whitfield and the other plaintiffs in the case of Whitfield, et. al. vs. State Democratic Party, the Arkansas law regulating primary elections bears a significant share of the responsibility for the lack of black political power. The law requires that a candidate receive a "majority of all the votes cast for candidates for the office" in order to win the party's nomination. If no candidate wins a majority in the first round of primaries, then the two leading candidates face each other in a run-off election two weeks later. The plaintiffs alleged that the majority requirement deprived black voters in Phillips County of an equal opportunity to elect the candidates of their choice -- a violation of Section 2 of the Voting Rights Act of 1965 (as amended in 1982).

The cornerstone of their argument was the historical pattern of racially polarized voting in Phillips County: white voters vote exclusively for white candidates and black voters for black candidates. So if more than one white candidate sought a nomination, the white vote would be divided and the black candidate might win the support of a plurality by winning all the votes cast by black voters. But there would be no chance for a black candidate to win a majority of the votes for nomination. The majority requirement would, then, force a run-off two weeks later. And in the run-off, the whites would close ranks and defeat the black candidate.

In 1986 and 1988, for example, four blacks came in first as plurality winners in preferential primaries, only to be defeated by racially polarized voting in run-offs. In neither the first or second primary did any white person publicly support or endorse a black candidate. In fact, Rev. Julius McGruder, black political candidate and former school board member, testified on the basis of 15 years working in elections that "no white candidate or white person has came out and supported no black." When Sam Whitfield won a first primary and requested support from Kenneth Stoner, a white candidate he defeated in the first round, Stoner told him that "He could not support a black man. He lives in this town. He is a farmer. His wife teaches school here and that there is just no way that he could support a black candidate."

But racially polarized voting is only one of the political disadvantages for blacks in Phillips County. Blacks also suffer disproportionately from poverty, and that poverty works to impede their effective participation in the run-off primary. For example, 42 percent of blacks lack any vehicle, while only nine percent of the white population are similarly handicapped; and 30 percent of blacks -- compared to eleven percent of whites -- have no telephone. Isolated by this poverty, black voters are less able to maneuver around such obstacles as frequent, last minute changes in polling places -- moving them to locations up to twelve to fifteen miles away, over dirt and gravel roads. Moreover, because of the number of blacks without cars, the lack of public transportation in the county, and the expense of taxis, the run-off election campaigns of black candidates must include "a get-out-and-vote kind of funding effort where we try to have cars to drive people to the polls."

Getting people to the polls a second time within a two week period severely limits the resources of black candidates, who have difficulty raising money a second time, paying for advertisements, notifying their supporters of the run-off election, and then convincing their supporters to go back to the polls a second time. Rev. McGruder testified that the run-off "just kill[s] all of the momentum, all of the hope, all of the faith, the belief in the system." According to McGruder, many voters "really can't understand the situation where you say `You know, Brother Whitfield won last night' and then come up to a grandma or my uncle, auntie and say `Hey, you know, we're going to have to run again in the next 10 days and -- because we've got a run-off.'"

In fact, between the first and second primary, turn-out drops precipitously, so that the so-called majority winner in the run-off may receive fewer votes than the black plurality winner in the first primary. For example, 1,893 fewer people voted in the run-off for County Judge in 1986 than in the first primary; and 1,725 fewer people voted in the 1986 circuit clerk run-off. In fact, in all three black/white run-off contests in 1986, the white run-off victor's majority occurred only because the number of people who came out to vote in the second primary went down.

The District Court that heard the 1988 challenge to the Arkansas law did not dispute the facts: that no black candidate had ever been elected to county-wide or state legislative office from Phillips County, and that "race has frequently dominated over qualifications and issues" in elections. Nevertheless, the court rejected the challenge to the majority requirement.

In the first place, the Court argued that the run-off requirement could not itself be blamed for the dilution of black voting strength. To be sure, bloc-voting by the white majority consistently prevented a relatively cohesive black population from nominating or electing their chosen representatives. But, the Court argued, that problem could not be solved by eliminating the run-off. On the contrary, if white candidates were stripped of the protection provided by the run-off, they would simply limit their numbers in the first round, self-selecting one white to run head-to-head against a black. So eliminating the run-off would "tend to perpetuate racial polarization and bloc-voting."

More fundamentally, however, the Court's decision was based more on its enthusiasm for the majority vote requirement than on its skepticism about the benefits of removing that requirement. Majority rule lies at "the very heart of our political system;" the requirement in the primaries was "not tenuous but, to the contrary, strong, laudable, reasonable, and fair to all." For a court to invalidate a majority vote requirement would undermine the operation of democratic systems of representation because "Americans have traditionally been schooled in the notion of majority rule. . . . [A] majority vote gives validation and credibility and invites acceptance; a plurality vote tends to lead to a lack of acceptance and instability."

The central place of majoritarianism in the Court's perception of the case can be highlighted with one more piece of background. Courts do, of course, sometimes invalidate voting schemes on the ground that those schemes deny a minority "an equal opportunity" to nominate or elect "candidates of their choice." In such cases, the court aims to ensure that all groups have some opportunity to have their interests represented in the governing body. For example, when an existing district has a black minority the standard remedy is to establish a subdistrict with a black majority in which blacks can elect representatives of their choice. By creating pockets in which minorities are majorities, race conscious districting provides a remedy for underrepresentation that respects the concerns of minorities while affirming the dominance of the majority principle.

But there was a problem with the standard remedy in Phillips County: the majority vote requirement applied to elections for seven county-wide positions. So each position represented the entire electorate. In such a circumstance, several courts have said that the statute does not apply. There can be no equal opportunity to elect because there is "no share" of a single-person office. Modifying an electoral structure to create alternative subdistrict majorities is not plausible where the majority vote rule applies to single-person offices.

With subdistricting ruled out, the only possible remedy then, would have been for the court to require the replacement of the majority requirement with a plurality system. And that is precisely what the plaintiffs asked the Court to do. But -- and here we return to the main point -- the Court would not require a plurality scheme because of its own conception of the central place of majoritarianism in democracy. While a plurality win in many cases is quite conventional, to order it as a remedy opened up possibilities of non-majoritarianism that the Court found quite threatening. With a plurality system unacceptable, and subdistricting unavailable, the court had no remedy for the plaintiffs. Absent a remedy, the Court found no violation.

At bottom then, the case was about democracy and majority rule. The Court denied relief because it identified democracy with majority rule. In fact, the Court actually suggested that if plaintiffs lost an election, their interpretation of the Act would require the Court simply to suspend elections. Where the Court saw democracy, however, the plaintiffs saw rule by a white numerical majority. As a numerical and stigmatized minority, they regarded the majority vote requirement as simply a white tool to "steal the election" -- a tool that had the effect of demobilizing black political participation, enhancing polarization rather than fostering debate and, in general, excluding black interests from the political process.

In short, the Court's conclusions were supported less by the evidence in Phillips County than by the Court's own majoritarian conception of democracy. To win the case then, the plaintiffs needed directly to challenge this premise about the intimate link between majority rule and democracy.

For example, they might have argued that majority rule is legitimate only when it is fair, and not simply because it is desirable to make decisions whose supporters outnumber their opponents. The conventional case for the fairness of majority rule, however, is that it is not really the rule of a fixed group -- The Majority -- on all issues; instead it is the rule of shifting majorities, as the losers at one time or one issue join with others and become part of the governing coalition at another time or another issue. So the argument for the majority principle connects it with the value of reciprocity: you cooperate when you lose in part because members of the current majority will cooperate when you win and they lose. The result will be a fair system of mutually beneficial cooperation.

But when a prejudiced majority excludes, refuses to inform itself about, or even seeks to thwart the preferences of the minority, then majority rule loses its link with the ideal of reciprocity, and so its moral authority. As the plaintiffs' evidence conclusively demonstrated, this was precisely the situation in Phillips County, where the fairness of the majority requirement was destroyed by the extreme racial polarization, the absence of reciprocity, and the artificial majorities created in the run-offs.

Racial Districting

Under conditions of sharp racial division then, majority rule can serve as an instrument to suppress a minority. It is not a fair way to resolve disagreements because it no longer promises reciprocity. That is what we learn from Brother Rice High School and Phillips County, Arkansas.

How can this unfairness be remedied? Perhaps through a more vigorous application of the conventional remedy of race-conscious districting. As indicated earlier, this remedy is not applied when -- as in Phillips County -- there is a system of district-wide, single-person offices rather than a collective decision-making body with multiple seats. But in the face of evidence of racial subordination, courts could simply reject such arrangements, require a system of subdistricts, and then ensure that some subdistricts have, for example, a black majority.

This strategy of judicially imposing alternative electoral constituencies is in fact consistent with some understandings of the Voting Rights Act. But instead of dwelling on legal issues, since race-conscious districting is, like any system of winner-take-all, flawed as a method for ensuring a fair system of political representation -- I'll argue and evaluate the conventional remedy on three tests of political fairness from the perspective of minority interests: Does a system mobilize or discourage participation? Does it encourage genuine debate or foster polarization? Does it promise real inclusion or only token representation? Race-conscious districting does not do well on any of these three dimensions. While it may be true that no election structure alone can do all that I envision, we need to consider alternatives to single-member districts -- in particular, to consider systems of proportional representation, which promise politically cohesive minorities both potential electoral success and proportionate influence throughout the extended political process. But I'm jumping ahead. First we need to see the problems with establishing geographically defined, race-conscious districts, each of which elects a single member to a representative body.

Firstly, districting fails to mobilize sustained voter participation. Districting systems rely on geography as a tool for identifying voter interests. Voters are assigned to territorial constituencies, the assumption being that territorial contiguity is associated with a community of interest.

Moreover, districting assumes that smaller political units best fulfill the political empowerment and equality norms of American constitutional jurisprudence. With smaller units, it's easier for constituents to identify political leaders, easier to recognize communities of interest, and easier to participate. Also, safe black districts enable blacks to get elected and then re-elected leading to positions of seniority and status within the elected body.

At the same time, however, single member districts emphasize individual candidacies, and for this reason may demobilize poor black and Latino voters. Turn-out generally goes up in response to first time election opportunities for the candidate of choice of the minority community. But the mobilization efforts for these break-through elections are generally not repeated. As the black pioneer becomes the incumbent, turn-out drops.

Moreover, politics in geographically organized, winner-take all electoral districts, tend to develop an exclusively electoral focus. The result is that the core constituency becomes alienated given the absence of local, alternative community organizations to educate and mobilize constituent participation outside of elections. Indeed, once elected, incumbents may demobilize constituents by not maintaining a genuine, community-based political organization to provide feedback, ideas, and reinforcement to the elected official while in office.

Furthermore, given racially-polarized voting, single-member districts give rise to real gains only to the extent there is substantial residential segregation at the appropriate geographic scale. Thus, for Latinos who live in dispersed barrios, districting does not capture either their real or potential power. In addition, it may exacerbate intergroup conflict as minorities are pitted against each other in a fight to be "the group" who gets the district. In jurisdictions with a complicated racial, ethnic, and linguistic mix, the redistricting struggle often becomes a source of conflict between blacks and other minority voters. These groups may compete over how many minority districts should be created and who should control them. Each group is encouraged to assert its superior moral, historic, and pragmatic claim. But in the lottery of competing oppression, no one wins.

Single-member districts also tend to under-represent minority voters, even where some race-conscious districts are created. Unless minority voters are both large enough and concentrated "just right," they will not enjoy representation in proportion to their presence in the population. Thus after a decade of race-conscious districting, blacks are not proportionately represented in any of the Southern legislatures.

Districting then, is limited as a strategy of empowerment. It is also troubling because it denies the connection between empowerment and voluntary participation. Districts are drawn by professionals without any involvement by voters. The presumption is that their political interests can be identified by their residential choices. The process of creating districts, which may be accompanied by extensive but perfunctory public hearings, is dominated by incumbent self-interest or court-appointed experts with no particular tie to grass roots concerns. By removing the issue from the voter, the districting process is antithetical to empowerment strategies based on voter participation and voter choice.

Furthermore, districting assumes that even where voters' interests are not represented by the ultimate district winner, their real interests will somehow, indirectly get represented. Consider a earlier Supreme Court case, UJO v. Carey. Hasidic Jews in Brooklyn challenged a legislative plan on the grounds that it discriminated against them by employing race-conscious districting. The Court denied the claim, holding that the challenged plan conveyed no stigma or disadvantage and so was not unconstitutional. In arriving at this conclusion, the Court in effect treated the Hasidim as just another group of white voters. The Court argued that white voters were proportionately represented state-wide, so drawing districts to represent blacks did not disadvantage whites -- even a white, Hasidic minority within a majority black district. This argument assumes that voters are interchangeable, that the courts can override voters' self-categorizations, and that the courts know better than the voters when their real interests are represented.

Secondly, districting fails to foster genuine debate about issues. In geographic districts, the threshold of representation is put close to 50 percent because districts are winner-take-all. The idea is that the winning candidate must demonstrate significant support to justify gaining all the power, but this pushes candidates toward the middle of the political spectrum where most of the votes are. The focus on developing consensus prior to the election means that issues are often not fully articulated or debated. Candidates avoid controversial positions and instead offer palliatives designed to offend no one.

Furthermore, the fact that interests within the district are only represented to the extent they garner majority support dooms third political parties to perpetual defeat. Because parties only win anything if they win a majority, and because it is difficult for a third party to win a majority, voters are reluctant to "throw away" their votes on third party candidates. And this in turn narrows the scope of political debate.

By promoting only two real choices, winner-take-all districting also has created an environment conducive to negative campaigning. A system that fosters two-way races is the "basis for such tried-and-true strategies as driving up the negatives of [an opponent] without worrying that the defecting voters will turn to a relatively unsullied third candidate," a common occurrence in three-way races. Negative campaign tactics contribute to voter alienation and apathy.

And thirdly, districting fails to promote opportunities for more than token inclusion. Proponents of race-conscious districting believe that separate can be made equal, or at least that poor blacks are empowered when provided a choice to elect a representative accountable only to them. Thus, they assume that electoral control works as a proxy for interest.

Creating majority black, predominantly poor districts is one way to ensure at least physical representation of black interests. This makes sense to the extent that electoral control insures accountability and influence. But racial districting also means that the electoral success of white legislators in white districts is not dependent on black votes. The direct consequence of majority black districts is that fewer white legislators are directly accountable to black interests. In this way, districting may reproduce within the legislature the polarization experienced at the polls; token electoral presence is replaced by token legislative presence.

Where blacks and whites are geographically separate, race-conscious districting isolates blacks from potential white allies -- for example, white women -- who are not geographically concentrated. It "wastes" the votes of white liberals who may be submerged within white, Republican districts. Thus it suppresses the potential development of issue-based campaigning and cross-racial coalitions.

As a consequence, race-conscious districting does not give blacks proportional legislative influence. Because majority black districts are necessarily accompanied by majority white districts, black representatives may be isolated in the governing body. Some conservative critics of race-conscious districting argue that such districts quarantine poor blacks in inner-city ghettoes. For example, critics of newly-drawn majority black congressional districts claim that the districts ultimately benefit white Republicans. In Alabama, for example, one critic claims such a district "yokes together by violence" areas as geographically different but racially similar as a large section of industrial Birmingham and vast expanses of the rural Black Belt, thus leaving other districts whiter and more Republican.

What these and related difficulties underscore is that proportionate influence requires something that winner-take-all districts simply cannot provide to numerically-weak minority voters: a basis of inclusion and representation that does not require winning more than 50 percent of the votes in a politically-imposed and geographically-defined constituency.


Let's now consider a system of proportional representation that drops the majority requirement. There are many such systems, but here I will focus on a scheme used in corporate governance called "cumulative voting." Under cumulative voting, voters cast multiple votes up to the number fixed by the number of open seats. If there are five seats on the city council, then each voter gets to cast five votes; if there are thirty songs at the senior prom, then each senior gets thirty votes. But they may choose to express the intensity of their preferences by concentrating all of their votes on a single candidate or a single song.

Let's return now to the three tests sketched earlier, and consider how cumulative voting fares in mobilizing participation, encouraging debate, and fostering inclusion.

Cumulative Voting and Participation. If voting is polarized along racial lines, as voting rights litigation cases hypothesize, then a system of cumulative voting would likely operate to provide at least a minimal level of minority representation. Unlike race-conscious districting, however, cumulative voting allows minority group members to identify their own allegiances and their preferences based on their strategic use of multiple voting possibilities. Instead of having the government authoritatively assign people to groups and districts, cumulative voting allows voluntary interest constituencies to form and regroup at each election; voters in effect "redistrict" themselves at every election. By abandoning geographic districting, it also permits a fair representation of minority voters who do not enjoy the numerical strength to become a district electoral majority or who -- as in my earlier example of Latinos living in dispersed barrios -- are so geographically separated within a large metropolitan area that their strength cannot be maximized within one or more single-member districts.

In all of these ways, cumulative voting would likely encourage greater electoral participation.

Cumulative Voting and Political Debate. Cumulative voting also looks good as a way to encourage genuine debate rather than foster polarization. Cumulative voting lowers the barriers to entry for third parties since supporters of such parties can concentrate all their votes on the candidates from their party. With those barriers reduced, minority political parties might reclaim, at a newly invigorated grassroots level, the traditional party role of mobilizing voter participation, expanding the space of organized alternatives, and so stretching the limits of political debate. Additionally, locally-based political parties might then organize around issues or issue-based coalitions. Since the potential support for the minority political party is not confined by a geographic or necessarily racial base, cross-racial coalitions are possible.

Cumulative Voting and Inclusion. Cumulative voting is more inclusive than winner-take-all race-conscious districting. Cumulative voting begins with the proposition that a consensus model of power sharing is preferable to a majoritarian model of centralized, winner-take-all accountability and popular sovereignty. It takes the idea of democracy by consensus and compromise and structures it in a deliberative, collective decision-making body in which the prejudiced white majority is "disaggregated." The majority is disaggregated both because the threshold for participation and representation is lowered to something less than 51 percent and because minorities are not simply shunted into "their own districts." These changes would encourage and reward efforts to build electoral alliances with minorities.

To get the full benefits of cumulative voting, however, it would also be necessary to change the process of governmental decision-making itself, away from a majoritarian model toward one of proportional power. In particular, efforts to centralize authority in a single executive would be discouraged in favor of power-sharing alternatives that emphasize collective decision-making. Within the legislature, rules would be preferred that require super-majorities for the enactment of certain decisions so that minority groups have an effective veto, thus forcing the majority to bargain with them and include them in any winning coalition, and with other devices for minority incorporation such as rotation in legislative office. Other electoral and legislative decision-making alternatives also exist such as legislative cumulative voting, that are fair and legitimate and that preserve representational authenticity, yet are more likely than current practices to promote just results.

The principle of proportionality is molded by the hope that a more cooperative political style of deliberation and ultimately a more equal basis for preference satisfaction is possible when authentic minority representatives are reinforced by structures to empower them at every stage of the political process. Ultimately however, representation and participation based on principles of proportionality are also an attempt to reconceptualize the ideal of political equality, and so the ideal of democracy itself.

The aim of that reconstruction should be to re-orient our political imagination away from the chimera of achieving a physically integrated legislature in a color-blind society and toward a clearer vision of a fair and just society. In the debate over competing claims to democratic legitimacy based on the value of minority group representation, I side with the advocates of an integrated, diverse legislature. A homogeneous legislature in a heterogeneous society is simply not legitimate.

But while black legislative visibility is an important measure of electoral fairness, taken by itself it represents an anemic approach to political fairness and justice. A vision of fairness and justice must begin to imagine a full and effective voice for disadvantaged minorities, a voice that is accountable to self-identified community interests, a voice that persuades, and a voice that is included in and resonates throughout the political process. That voice will not be achieved by majoritarian means or by enforced separation into winner-take-all racial districts. For in the end democracy is not about rule by the powerful -- even a powerful majority -- nor is it about arbitrarily separating groups to create separate majorities in order to increase their share. Instead, the ideal of democracy promises a fair discussion among self-defined equals about how to achieve our common aspirations. To redeem that promise, we need to put the idea of proportionality at the center of our conception of representation.

Originally published in the September-October 1992 issue of Boston Review


Copyright Boston Review, 1993–2005. All rights reserved. Please do not reproduce without permission.

 | home | new democracy forum | fiction, film, poetry | archives | masthead | subscribe |