I. Alternative Electoral Systems As The Solution To A Statutory-Constitutional Dilemma

Steven J. Mulroy

A. Vote Dilution and Single-Member Districts

Congress passed the Voting Rights Act in 1965 to address the problem of racial discrimination in voting. When enforcement of the Act in the 1960s and 1970s led to increased voter registration among formerly excluded minority groups,[11] voting rights advocates discovered still another hurdle to effective participation by these groups in American democracy: minority vote dilution. Simply put, electoral systems were in many places designed to prevent minority voters from exerting any real influence in electoral outcomes, thanks to the bloc voting patterns of the white majority. This vote dilution was accomplished in two ways: through the use of at-large electoral systems, and through districts ěgerrymanderedî to minimize minority voting strength. <<PK: I donít think a fn necessary here--I develop these concepts and cite where appropriate in the paragraphs above.>>

The traditional, ěwinner-take-allî form of at-large election was (and is) commonly used for local elections such as for city councils or county commissions. <<PK: I donít think fn necessary here.>>It has voters from all over the jurisdiction voting to fill open seats on jurisdiction-wide legislative bodies. Under this method of election, each voter may cast only one vote for each candidate for a particular office, up to the number of empty seats, with the top vote-getters filling these seats. A related approach elects several representatives each from large ěmultimemberî districts. Within each such multimember district, elections operate in a manner similar to at-large elections. Most local jurisdictions in the United States use single-member districts, winner-take-all at-large elections, or some combination of the two.[12]

It has long been acknowledged that where voting is polarized along racial or ethnic lines, these ětraditionalî or ěwinner-take-allî at-large/multimember methods allow the white majority to vote as a bloc and fill all seats, ensuring the defeat of candidates backed by minority voters and thus ědilutingî minority voting strength.[13] For example, suppose a five-member county commission in a county where black persons[14] constitute 40% of the population. Where, as is commonly the case, black voters consistently prefer different candidates from white voters, the 3/5 white majority can continually ěshut outî the black minority and fill all five seats with white-preferred candidates. Even though the politically cohesive voting bloc of black voters represents 2/5 of the populace, they are unable to elect a single candidate of choice.[15] This inability is the arbitrary result of the particular method of election interacting with the prejudices of white voters.

Minority vote dilution also occurs in settings using single-member districts, geographic electoral units from which only a single candidate is elected, as in the U.S. House of Representatives. Just as these districts can be politically gerrymandered to disproportionately elect Democrats or Republicans, they can be drawn to disproportionately elect white-preferred candidates and underrepresent minority-preferred candidates.

The canonical remedy for minority vote dilution through electoral method (i.e., through at-large/multimember systems) and through gerrymandered districting has been to draw single-member districts with significant minority populations.[16] Litigants can force the creation of such minority-oriented districts through lawsuits under Section 2 of the Voting Rights Act.

B. Section 2 and Gingles

The statutory solution to this problem is set out in Section 2 of the Voting Rights Act, which Congress passed in 1965. Section 2 prohibits States and political subdivisions from using any standards, practices, or procedures that abridge the right to vote of any member of a protected class of racial and ethnic minorities.[17] In 1982, Congress amended Section 2 to clarify that it intended to apply a ěresults testî to Section 2 vote dilution claims. Rather than requiring a showing of discriminatory purpose on the part of the government body implementing the challenged electoral practice, a plaintiff could prevail simply by showing that the ětotality of circumstancesî reveal that ěthe political processes leading to nomination or election ... are not equally open to participationî by members of the protected class, because those members ěhave less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.î[18] The legislative history of the amendment sets out a lengthy, nonexhaustive list of factors to be considered under the ětotality of circumstances,î[19] referred to commonly as the ěSenate factors.î

In 1986, the Supreme Court clarified how courts were to implement the ěresults testî amendments in Thornburg v. Gingles, a Section 2 challenge to certain North Carolina state legislative multimember districts.[20] In Gingles, the Court established that Section 2 plaintiffs must establish three ěpreconditionsî to make out a prima facie case: (1) the minority group is sufficiently numerous and compact to form a majority in a single-member district; (2) the minority group is politically cohesive (i.e., its members tend to vote alike); and (3) the white majority votes sufficiently as a bloc so as usually to defeat the minorityís preferred candidate.[21] The first of these preconditions is often referred to simply as ěcompactness,î and the second and third are often referred to together as ěracial bloc votingî or ěracially polarized voting.î[22] If plaintiffs failed to prove any of these three preconditions, they lost. If they could establish all three, they still had to prove a violation under the ětotality of the circumstancesî analysis, which incorporated the ěSenate factors.î[23]

For years before Gingles, and with even more vigor for seven years thereafter, voting rights litigants used Section 2 to force the creation of numerous minority-oriented districts at the congressional, state, and local level.[24] Some, but not all, of these districts were oddly-shaped. Sometimes the districts were oddly shaped because that was the only way to create majority-minority districts while still satisfying one-person, one-vote concerns.[25] At other times, the odd shapes were the result of incumbent legislators trying to satisfy the Voting Rights Act while simultaneously protecting their own incumbency.[26] The minority districts were often no more oddly shaped than other, non-minority districts drawn by the same jurisdiction.[27] The degree of non-compactness of (both minority and non-minority) districts increased over time as improvements in redistricting computer technology allowed ever-more-sophisticated fine-tuning of demographic data by drawers of redistricting plans.[28] Nonetheless, the race-influenced oddities were bound to attract the attention of the Supreme Court, and in 1993 one such district did.

C. Shaw, Miller, and Reverse-Discrimination "Racial Gerrymanders"

In Shaw v. Reno,[29] the Supreme Court considered a challenge to North Carolinaís Twelfth congressional district, a long, snake-like minority-majority district which stretched along Interstate 85.[30] In Shaw, the Court recognized a new constitutional cause of action challenging an oddly-shaped minority district as a ěracial gerrymanderî under the Fourteenth Amendment.[31] This new cause of action was ěanalytically distinctî from the traditional vote dilution claim and did not require proof that a particular groupís vote was in fact diluted.[32] Instead, the harm caused by such ěgerrymanderingî was two-fold: it reinforced the stereotype that members of the same racial group think and vote alike (ěstigmatic harmî), and it encouraged elected officials from such districts to represent only the members of their racial group rather than the district as a whole (ěrepresentational harmî).[33]

As explained in succeeding cases, a Shaw plaintiff confronts a threshold question as to whether race is the ěpredominant factorî driving the configuration of the challenged minority district.[34] To make this showing, the plaintiff must prove that the challenged district ěsubordinated [to race] traditional race-neutral districting principles,î such as ěcompactness, contiguity, respect for political subdivisions, or communities defined by actual shared interests.î[35] If the plaintiff fails to make this showing, the minority district will be upheld and the analysis ends there. If the plaintiff meets the ěpredominant factorî threshold, the constitutional standard of ěstrict scrutinyî applies: the district will be invalidated unless its configuration was narrowly tailored to serve a compelling governmental interest.[36] Although the Court has not detailed a definitive list of governmental interests which would support a Shaw-type racial gerrymander, compliance with Section 2 is likely one such interest.[37]

The Shaw line of cases has severely curtailed the ability of legislators, litigants, and courts to create single-member districts which provide minority voters with a realistic opportunity to elect candidates of choice.[38] In doing so, the cases create a dilemma for anyone interested in remedying minority vote dilution. Persons interested in assuring fair representation to minorities must steer a narrow middle course between the Scylla of the Voting Rights Act, which requires that race be taken into account when drawing districts, and the Charybdis of Shaw, which requires that race not be used ětoo much.î The performers in the upcoming 2000 redistricting circus will be walking a tightrope, and no net can cushion them from the risk of protracted, expensive litigation and disrupted election cycles. These pitfalls can be avoided, however, by eschewing districting altogether and adopting alternative electoral systems.[39]

D. Alternative Electoral Systems

Although at-large systems used in the United States have traditionally been of the ěwinner-take-allî variety, other varieties have been and are employed here.[40] These systems employ special voting rules designed to enhance the ability of voting minorities to obtain some representation. There are three such systems which have a history of use in the United States. They can be used in partisan or non-partisan elections, and which have been discussed as non-district alternative voting rights remedies at the local level.[41] These three systems are limited voting, cumulative voting, and preference voting (also known as choice voting or the Single Transferable Vote).[42]

In a limited voting system, a voter casts one vote per candidate to fill a number of seats, but the total number of votes a voter may cast is lower than the total number of seats to be filled.[43] For example, a voter might be allowed to vote for only four candidates to fill five seats on a city council. This limitation is designed to prevent the majority "from making a clean sweep of all seats by voting a straight ticket."[44] Limited voting systems are used in numerous local jurisdictions in North Carolina, Alabama, Connecticut and Pennsylvania. [45]

In a cumulative voting system, voters have X number of votes and can distribute them among the candidates any way they see fit. This number of votes is usually, but need not be, equal to the number of seats to be filled.[46] If a voter has five votes, she can cast one vote each for her top five preferred candidates, she can "plump" all five votes for one candidate for whom she has an especially strong preference, or she can divide her votes 3-2 between her top two choices, and so on. [47] Cumulative voting is used in scores of jurisdictions in Texas and Alabama, as well as in Peoria, Illinois.[48] Historically, it was used to elect members of the Illinois House of Representatives from 1870 to 1980.[49]

In preference voting,[50] the voter ranks the candidates in her order of preference, putting beside each candidate's name a "1," "2," etc., for as many candidates as she wishes. The votes are counted in a series of rounds. In the first round, candidates netting more than a certain minimum amount of first-choice votes [51] are assigned a seat; their "excess" votes (i.e., the number of votes they received above the minimum) are reassigned to other candidates based on the votersí second-choice selections. In the second round, any candidates above the minimum quota of first-choice votes (after reassignment) are seated; if there are no such candidates, the lowest vote-getter is disqualified, and his votes are reassigned on the same second-choice basis. This process of seating and disqualifying candidates, and reassigning their votes accordingly, continues through as many iterations as necessary to fill all the seats.[52] Preference voting is used in Ireland and Australia, among other countries.[53] In the United States, it is used in city council and school board elections in Cambridge, Massachusetts, and in local community school board elections in New York City.[54] Historically, it was used to elect city councils in approximately two dozen American cities throughout all regions of the United States.[55]

For each of these three alternative systems, a mathematical formula exists which determines the minimum percentage of the electorate a politically cohesive minority group needs to be in order be assured of electing at least one candidate of choice. This formula, universally acknowledged by political scientists, is called the ěthreshold of exclusion.î For both cumulative and preference voting,[56] the threshold of exclusion is:

1 / (# seats to be filled) + 1 [57]

These alternative systems have proven to be effective at enhancing the ability of racial and ethnic minority voters to elect candidates of choice.[58] Importantly, they do so in a race-neutral manner, and thus do not raise Shaw concerns.[59] For this reason, commentators[60] have argued that courts should use these alternative systems as remedies in minority vote dilution lawsuits under the Voting Rights Act. In many cases, such remedies have been adopted by consent decree in Section 2 litigation,[61] but they have rarely been imposed by judicial fiat as a Section 2 remedy over the objection of a defendant jurisdiction.


Part Two, Authority And Standards For ěThe Way Outî Authority of Courts to Adopt

[11] Evidence regarding the increase of minority voter registration in the years following the passage of the Voting Rights Act can be found in James E. Alt, The Impact of the Voting Rights Act on Black and White Voter Registration in the South, in Quiet Revolution in the South: The Impact of the Voting Rights Act, 1965‑1990 at 351, 352-354 (1994) (summarizing findings); see also id. at 351‑73.

[12] See Edward Still, Alternatives to Single-Member Districts, in Minority Vote Dilution Alternatives to Single-Member Districts, in 249, 249 (Chandler Davidson ed., 1989). The dominance of these systems is a historical product, stemming from their use in England.

[13] See Rogers v. Lodge, 458 U.S. 613, 616 (1982); Chapman v. Meier, 420 U.S. 1, 17-18 (1975).

[14] For much of this article, I use black persons in examples involving minority groups suffering vote dilution. These examples should be understood to apply equally to members of any minority group protected under the Voting Rights Act.

[15] This was exactly the situation in Granville County, North Carolina, site of a vote dilution lawsuit under the Voting Rights Act. See McGhee v. Granville County, 860 F.2d 110 (4th Cir. 1988). Situations like this have been and continue to be common. e.g., Teague v. Attala County, 92 F.3d 283, 283 (5th Cir. 1996) (no blacks elected at-large in modern times despite blacks making up 40% of population).

[16] See, e.g., Major v. Treen, 574 F. Supp. 325, 355 (E.D. La. 1983) (ordering creation of black-majority congressional district in New Orleans); Westwego Citizens for Better Government v. City of Westwego, 946 F.2d 1109, 1124 (5th Cir. 1991) (ordering creation of remedial districting plan to remedy minority vote dilution).

[17] The Voting Rights Act bars voting-related discrimination against persons on account of race, color, national origin, and membership in a ělanguage minority group.î 42 U.S.C. ß 1973, 1973b(f) (1994). The Act defines a ělanguage minority groupî to include Asian-Americans, Hispanics, Native Americans, and Alaskan Natives. 42 U.S.C. ß 1973l(c) (1994).

[18] 42 U.S.C. ß 1973(b).

[19] The Senate Judiciary Committee majority report on the Act lists the following factors, focused on the particular defendants State or political subdivision: (1) the history of official discrimination affecting the right to vote; (2) the degree to which voting was racially polarized; (3) the use of other dilutive voting procedures such as majority vote requirements; (4) any denial of minority candidate access to candidate slating processes; (5) socioeconomic disparities on the part of the minority group members; (6) racial appeals in campaigns; (7) the degree of minority candidate electoral success; (8) the degree of responsiveness on the part of elected officials to the concerns of the minority group; and (9) the extent to which the policy underlying the challenged practice or procedure is tenuous. See S. Rep. No. 97-417, at 28-29 (1982), reprinted in 1982 U.S.C.C.A.N. 205-207.

[20] 478 U.S. 30, 30 (1986).

[21] Id. at 50-51.

[22] Id. at 55; Teague v. Attala County, 92 F.3d 283, 287 (5th Cir. 1996).

[23] Id. at 43-52.

[24] See, e.g., Major v. Treen, 574 F. Supp. 325, 355 (E.D. La. 1983) (ordering creation of black-majority congressional district in New Orleans); DeGrandy v. Wetherell, 815 F. Supp. 1550, 1560 (N.D. Fla. 1992) (invalidating Florida state legislative districting plan), in part, reversed in part, Johnson v. De Grandy, 512 U.S. 997 (1994); Westwego Citizens for Better Government v. City of Westwego, 946 F.2d 1109, 1124 (5th Cir. 1991) (ordering creation of local remedial districting plan to remedy minority vote dilution).

[25] See Reynolds v. Sims, 377 U.S. 533, 559 (1964) (explaining one-person, one-vote requirement); see, e.g., In Re Constitutionality of SJR 2-G, 601 So.2d 543, 546 (Fla. 1992) (explaining odd shape of state senate district in this manner).

[26] See, e.g., Shaw v. Reno, 509 U.S. 630, 672 & n.10 (1993) (noting assertions that odd shape of congressional district can be explained in this manner).

[27] See, e.g., Lawyer v. United States, 117 S.Ct. 2186, 2194-95 (1997) (upholding constitutional challenge to minority state senate district in part on this ground); Miller v. Johnson, 515 U.S. 900, 940 (1995) (describing evidence to this effect regarding Georgia minority congressional district) (Ginsburg, J., dissenting).

[28] See Vera v. Bush, 116 S.Ct. 1941, 1953 (1995) (emphasizing evidence of advanced redistricting computer abilities) (opinion of OíConnor, J.); Declaration of John Guthrie, Scott v. United States Dept. Of Justice, No. 94-622-Civ-T-23C (M.D. Fla.) (three-judge court) (filed Nov. 17, 1995) (describing evolution of redistricting computer technology and its resulting increase in non-compact districts).

[29] 509 U.S. 630 (1993).

[30] Id. at 630.

[31] Id. at 657.

[32] Id. at 650-653. That is, even if a Shaw plaintiff could point to no cohesive group which would have less of an opportunity to elect candidates of choice as a result of the challenged minority district, or which would be significantly underrepresent ed in proportion to its share of the population, the plaintiff could still prevail if it made out the other elements of the Shaw claim.

[33] Id. at 647-648; see also United States v. Hays, 515 U.S. 737, 744 (1995) (further explaining stigmatic and representational harm).

[34] See Miller v. Johnson, 515 U.S. 900, 915-916 (1995); Vera v. Bush, 116 S.Ct. 1941, 1951 (1996).

[35] Id. at 916.

[36] Id.

[37] In Vera v. Bush, 116 S.Ct. 1941 (1996), supra, Justice OíConnor held that compliance with the ěresults testî of Section 2 could be a sufficiently compelling state interest to justify a ěracial gerrymanderî subject to strict scrutiny. 116 S.Ct. at 1968-1970. This is in accord with the views of four other Supreme Court justices. id. at 1974, 1989 (Stevens, J., dissenting, joined by Breyer & Ginsburg, J.J.); id. at 1997, 2007 (Souter, J., dissenting, joined by Breyer and Ginsburg, J.J.). The remaining Justices who have dealt with the issue recently have assumed this point without deciding it. See id. at 1960.

[38] Almost every Supreme Court opinion considering a Shaw challenge on the merits has invalidated the challenged district(s) and issued language cautioning against over-using race in making redistricting decisions. See Miller v. Johnson, 115 S.Ct. 2475, 2493-2494 (invalidating minority congressional district in Georgia); Bush v. Vera, 116 S.Ct. 1941, 1951 (1996) (invalidating three minority congressional districts in Texas); Shaw v. Hunt, 517 U.S. 899, 917 (1996) (Shaw II) (invalidating North Carolina congressional district); Abrams v. Johnson, 117 S.Ct. 1925, 1935 (1997) (criticizing other proposed minority congressional districts); see Lawyer v. Department of Justice, 117 S.Ct. 2186, 2195 (1997) (upholding minority state senate district in Florida). The repeated Supreme Court criticism has, to put it mildly, the potential for causing a chilling effect on the creation of minority-oriented single-member districts.

[39] By advocating this course, I do not mean to suggest agreement with the Shaw line of decisions, which I believe (for reasons outside the scope of this article) to have been wrongly decided. Although I consider alternative electoral methods often to be preferable to single-member districts, infra, I consider such districts generally preferable to traditional, dilutive, ěwinner-take-allî at-large methods of election, and I acknowledge that there are circumstances where minority districts are the best vote dilution remedy.

[40] There is no federal constitutional or statutory provision banning the use of alternative electoral schemes, or their adoption by federal courts. See Holder v. Hall, 512 U.S. 874, 896-99, 908-13 (1994) (Thomas, J., concurring); LULAC v. Clements, 986 F.2d 728, 814-15 (5th Cir. 1993), on other grounds, 999 F.2d 831 (5th Cir. 1993) (en banc); United States v. Marengo County Commín, 731 F.2d 1546, 1560 (11th Cir. 1984); Dillard v. Town of Louisville, 730 F. Supp. 1546, 1548 (M.D. Ala. 1990).

[41] See, e.g., Richard L. Engstrom, Multi-Seat Election Systems as Remedies for Minority Vote Dilution, 21 Stetson L. Rev. 743, 750 (1992) (hereinafter ěModified Multi-Seat Systemsî).

[42] The description of these three alternative systems which follows borrows from Mulroy, supra note 8. For a fuller discussion of these systems, see that article at 14-22.

Alternative electoral systems are often referred to as ěproportional representationî systems. This usage raises two definitional concerns. First, although all three systems tend to produce electoral results that are more ěproportionalî to voter preferences than winner-take-all systems, only preference voting produces results that are reliably ěproportionalî as a strictly defined, mathematical matter. Therefore, political scientists consider only preference voting (also known as "choice voting" or the ěsingle transferable voteî) an example of ěproportional representation,î calling limited and cumulative voting ěsemi-proportionalî systems. Douglas Amy, Real Choices, New Voices 186 (1993); Edward Still, Alternatives to Single-Member Districts, in Minority Vote Dilution 249, 258 (Chandler Davidson ed., 1989) . Second, these systems should not be confused with the party-list, parliamentary systems used in Europe. These systems are also examples of proportional representation, but have distinct features alien to the American experience, such as voting rules requiring voters to vote for parties, not candidates, and a weak executive whose administration may be truncated by a vote of those legislators. See Amy, supra, at 227-230. Such parliamentary features are not found in the alternative systems used in the United States, and I do not advocate their adoption here. See Mulroy, supra note 8, at 339.

[43] See Still, Alternatives to Single-Member Districts, supra note 42, at 249, 253.

[44] Id.

[45] See Cleveland County Assoc. For Government By The People v. Cleveland County Bd. of Commírs, 965 F. Supp. 72, 79 & n.9 (D.D.C. 1997) (describing use in three county and two municipal elected bodies in North Carolina), on other grounds, 142 F.3d 468 (D.C. Cir. 1998); Dillard v. Baldwin County Bd. of Educ., 686 F. Supp. 1459, 1460 (M.D. Ala. 1988) (describing the numerous Alabama counties and municipalities subject to the Dillard litigation, and tracing the litigation's history); Amy, Real Choices, New Voices, supra note 42 at 217, 232 (noting use in a "few cities and counties" in Connecticut and Pennsylvania).

[46] See Note, Remedies for Unlawful At-Large Systems, 92 Yale L. J. 144, 153 n.40 (1982) (hereinafter ěAlternative Voting Systemsî).

[47] The ability of voters to cast more than one vote is not a violation of the "one person, one vote" constitutional principle because each voter is entitled to the same number of votes. See Reynolds v. Sims, 377 U.S. 533, 559 (1964) (explaining that gravamen of the principle is that "one man's vote ... is to be worth as much as another's"); also Orloski v. Davis, 564 F. Supp. 526, 530 (M.D. Pa. 1983) (upholding limited voting scheme over one person, one vote objections); Lo Frisco v. Schaffer, 341 F. Supp. 743, 748 (D. Conn.) (same), aff'd, 409 U.S. 972 (1972); Kaelin v. Warden, 334 F. Supp. 602, 605 (E.D. Pa. 1971) (same). also Lani Guinier, The Triumph Of Tokenism: The Voting Rights Act and the Theory of Black Electoral Success, 89 Mich. L. Rev. 1088, 1148 n.332 (1991), and sources cited therein (explaining why these voting systems are consistent with the one person, one vote principle).

[48] See Robert Brischetto & Richard Engstrom, Voting And Latino Representation: Exit Surveys In Fifteen Different Texas Communities, 78 Social Science Quarterly Voting And Latino Representation: Exit Surveys In Fifteen Different Texas Communities, 78 973 (1997) (describing uses in Texas by at least 15 municipalities and 26 school boards since 1991); Richard Engstrom, et al., Limited And Cumulative Voting In Alabama: An Assessment After Two Rounds Of Elections, 6 National Political Science Review 180, 185 (1997) (describing Alabama uses in four municipalities plus the county commission and school board of a separate county since 1988); Richard Engstrom, Modified Multi-Seat Election Systems, supra note 42, at 757 (1992) (describing use in Peoria since 1991).

[49] Amy, Real Choices, New Voices, supra note 42, at 186.

[50] Preference voting is sometimes referred to as the "single transferrable vote"; or as the "Hare system," which is a particular methodology for counting preference vote ballots. Most recently, it has been called ěchoice voting.î

[51] This minimum is calculated by dividing the total number of votes cast by { (# of seats to be filled) + 1}, and then adding one. This is the so-called ěDroop quota.î Still, supra, at 259. Thus, where three seats are to be filled, the minimum is one more than 1/4 of the overall vote; where four seats are available, it is one more than 1/5 of the vote; and so on.

[52] See Still, Alternatives To Single-Member Districts, supra note 42, at 258-261. Good explanations of the counting process can be found in Amy, supra note 42, at 230-231 & Appendix C; and Engstrom, Modified Multi-Seat Election Systems, supra note 42, at 765-769. Engstrom provides a diagram that is particularly helpful.

[53] See Amy, supra note 42, at 18.

[54] See id. at 18, 137-138; Engstrom, Multi-Seat Election Systems, supra, note 42 at 766-767.

[55] See Engstrom, Modified Multi-Seat Election Systems, supra note 42, at 766-767. These cities adopted preference voting at various points from 1915 through 1950, and continued using it for periods varying from several years to several decades. See Leon Weaver, Rise, Decline, and Resurrection of Proportional Representation in Local Governments in the United States, in Electoral Laws and Their Political Consequences 139, 139-141 (B. Grofman & A. Lijphart eds., 1986); see also Robert A. Burnham, Reform, Politics, And Race In Cincinnati: Proportional Representation And The City Charter Committee, 1924-1959, 23 Journal Of Urban History 131, 132 (1997) (describing history of use in Ohio cities). New York City used preference voting to elect its City Council from 1936 to 1947. See Belle Zeller & Hugh A. Bone, Repeal Of Proportional Representation In New York City Ten Years In Retrospect, American Political Science Review [41] See, e.g., Richard L. Engstrom, Multi-Seat Election Systems as Remedies for Minority Vote Dilution in Modified Alternatives to Single-Member Districts,supra note 42, at 249, 253.

1127, 1127-1133 (1948). Repeal of preference voting in these cities often came as a result of public hostility to the election of blacks (as in Cincinnati) or Communists (as in New York). See Burnham, supra, at 153 (noting role of racial antagonism in repeal of PR in Cincinnati); Richard Engstrom, Cincinnatiís 1988 Proportional Representation Initiative, 9 Electoral Studies 217, 219-222 (1990) (same); Zeller & Bone, supra, at 1133 (noting role of fear of Communism in New York repeal); Amy, supra note 42, at 173 (same).

[56] For limited voting, the threshold of exclusion formula is:

(# of votes each voter has) / {(# of votes each voter has) + (# of seats to be filled)}.

Still, supra note 42, at 254; Engstrom, Modified Multi-Seat Systems, supra note 42, at 743, 758. Thus, if 5 seats are up for election, and each voter has 4 votes, the threshold is 4/(4+5) = 4/9 = 44% of the vote. Mathematically, the threshold is lower (and thus more accessible to minority voters) where the number of seats each voter has is limited to one. See Still, supra note 42, at 254. This can be seen by holding (# of seats to be filled) constant and plugging in increasing integers for (# of votes each voter has).

[57] Still, supra note 42, at 256.

[58] See infra Part II.D.1; also Mulroy, supra note 8, at 36-38 (discussing the effectiveness such systems have had in electing minority-preferred candidates).

[59] See Mulroy, supra note 8, at 38-51.

[60] See notes 3 and 4, supra.

[61] See for example, Judge Myron Thompsonís decisions approving cumulative voting remedies as settlements to several of the related Dillard cases brought under the Voting Rights Act. Dillard v. Baldwin County Bd. of Educ., 686 F. Supp. 1459, 1460 (M.D. Ala. 1988) (describing history of the Dillard litigation).

  Part Two, Authority And Standards For ěThe Way Outî Authority of Courts to Adopt