Summary
upholding plan for the election of county board of supervisors in which each town representative cast weighted number of votes that was directly proportional to his or her town's portion of the county's total population, in part because there was no suggestion that plan "contain[ed] any built-in bias favoring a particular political interest, race or geographic area"
Summary of this case from LARIOS v. COXOpinion
No. 94-CV-982.
May 15, 1995.
Jordan Walster, Herbert Jordan, Roxbury, NY, for plaintiffs.
Richard B. Spinney, Stamford, NY, for defendant.
MEMORANDUM-DECISION ORDER
In the case before the court, plaintiffs Roxbury Taxpayers Alliance and three voter-residents of Delaware County, challenge the system of apportionment which allocates weighted votes to members of the Delaware County Board of Supervisors. Plaintiffs allege that the system violates the one person, one vote principle implicit in the Equal Protection Clause of the 14th Amendment.
I.
A. The Parties
Plaintiff Roxbury Taxpayers Alliance is an unincorporated association concerned with voting rights and political representation in Delaware County, New York. Plaintiffs Susan E. Moore and Edward E. Moore are residents and registered voters of the Town of Middletown, Delaware County. Plaintiff Elsa McDonald is a resident and voter from the Town of Sidney, Delaware County, the most populous town in the county.
Defendant Delaware County Board of Supervisors is the county's chief legislative body. The Board of Supervisors was established pursuant to Article IX of the New York State Constitution and Article IV of the County Law of New York. It is responsible for adopting the weighted voting plan which is in use today and which is the subject of this suit.
B. Delaware County's Weighted Voting Plan
According to the 1990 census, Delaware County has a population of 47,225 and is divided into 19 townships ranging in population from 550 to 6667. In accordance with state and county laws, the registered voters in each town periodically elect one town supervisor. The elected town supervisor represents his or her town on the county Board of Supervisors. Thus, voters of each town are represented by a single member of the board, regardless of the population of the town in which they reside and vote.
Voting on the Board of Supervisors is governed by Local Law No. 4 of 1991. Under this law, votes are distributed according to each town's percentage of the total county population. For example, the Town of Bovina has a population of 550 out of a total county population of 47,225, and therefore constitutes 1.17% of the county population. Under the weighted voting plan, Bovina has 39 out of a total of 3480 votes, which represents 1.12% of the total votes. The Town of Sidney, population 6667, constitutes 14.12% of the county population and has 468 votes, representing 13.45% of the total number of votes.
This is the number of votes held by Bovina in the great majority of situations in which a simple majority vote is required. In the infrequent occasions when county laws require a supermajority of two-thirds and three-fifths to pass certain types of legislation, Bovina has 28 and 27 votes respectively. This constitutes 1.15% and 1.20% of the total votes.
The population of each town in Delaware County and the corresponding number of weighted votes is attached as Appendix 1. The percentage of population and corresponding percentage of weighted votes for each town is attached as Appendix 2.
These types of weighted voting is more prevalent in New York, where it is used by more than twenty counties, than anywhere else in the country. M. David Gefland Terry E. Allbritton, Conflict and Congruence in One-Person, One-Vote and Racial Dilution Litigation: Issues Resolved and Unresolved by Board of Estimate v. Morris, 6 J.L. Pol. 93, 112 (1989). Weighted voting was originally developed in response to the landmark case of Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). In Reynolds the Supreme Court found that apportionment plans in which districts of different sizes elected representatives who each had a single vote undervalued the franchise of voters in districts with populations above the mean, and impermissibly overvalued the franchise of voters in districts with populations below the mean. Id. at 563, 84 S.Ct. at 1382. The Reynolds Court therefore held such plans to be unconstitutional under the one person, one vote principle of the Equal Protection Clause of the 14th Amendment. Id. at 569, 84 S.Ct. at 1385.
By allocating votes in proportion to the population of the electoral district, weighted voting plans were used as a device for satisfying the one person one vote requirement, while simultaneously preserving traditional political subdivisions which often had wide variations in population.
For a historical analysis of the development of weighted voting plans, see M. David Gefland Terry A. Allbritton, Conflict and Congruence in One Person, One Vote and Racial Dilution Litigation: Issues Resolved and Unresolved by Board of Estimate v. Morris, 6 J.L. Pol. 93, 111-112 (1989).
C. Procedural Posture of the Case
The parties are in substantial agreement on all issues of material fact relating to the weighted voting system in Delaware County. Plaintiffs concede that the Delaware County weighted voting plan results in each board member casting votes "which [are] essentially proportionate to the population of his or her district." (Plaintiffs' Statement Pursuant to Local Rule 10(j), ¶ 9, dated Nov. 4, 1994). Plaintiffs do not, therefore, challenge the current system as it is applied to them. Instead, the Plaintiffs have moved for summary judgment under Fed.R.Civ.P. 56, on the ground that weighted voting has been held to be per se unconstitutional by the Supreme Court in Board of Estimate v. Morris, 489 U.S. 688, 109 S.Ct. 1433, 103 L.Ed.2d 717 (1989).
D. Standing
The Defendant correctly maintains that the Roxbury Taxpayers Alliance, Susan Moore and Edward Moore lack standing to bring this challenge. First, in League of Women Voters v. Nassau County Bd. of Supervisors, 737 F.2d 155 (2d Cir. 1984), the Second Circuit restricted organizational standing under § 1983 "by interpreting the rights [§ 1983] secures to be personal to those purportedly injured." Id. at 160 (citing Aguayo v. Richardson, 473 F.2d 1090 (2d Cir. 1973)). The Second Circuit concluded that "[n]either the language nor the history . . . [of § 1983] suggests that an organization may sue under the Civil Rights Act for the violations of rights of members." Id. Accordingly, the Roxbury Taxpayers Alliance, which is an association concerned with its members' voting rights and political representations, lacks standing to bring this action on behalf of its members.
Second, Plaintiffs Susan Moore and Edward Moore also lack standing, as they have failed to show injury under the weighted voting plan. In League of Women Voters v. Nassau County Bd. of Supervisors, the Second Circuit adopted the position of the Fifth Circuit in Fairley v. Patterson, 493 F.2d 598 (5th Cir. 1974), and held that plaintiffs who could not show injury by demonstrating that they were domiciled in an underrepresented voting district, "lacked standing to contest a reapportionment law." League of Women Voters, 737 F.2d at 161. Plaintiffs Susan Moore and Edward Moore are registered voters in the Town of Middletown which has 7.21% of the county population and 7.30% of the weighted votes. Because Middletown is slightly over represented under the current plan, Plaintiffs Susan Moore and Edward Moore can not show injury and therefore lack standing to challenge the apportionment plan.
While Middletown is slightly underrepresented in situations in which a two-thirds vote is required, with 7.17% of the weighted votes, the Town is over represented when a three-fifths vote is required, with 7.22% of the weighted votes. Thus, on balance, Middletown is slightly overrepresented.
The Defendant does not dispute the standing of Plaintiff MacDonald, who is a resident of Sidney. Sidney's percentage of the county's weighted votes is marginally lower than its percentage of the county population. As such, Plaintiff MacDonald has grounds for alleging injury and therefore has standing to prosecute this action.
II.
Turning now to the merits of the claim before the Court, it is based on the theory that "the Fourteenth Amendment requires that all local legislatures be structured so that the legislators represent districts which are at least roughly equal in population. A legislature which is not so structured cannot conform to the constitutional command by using weighted voting." (Pl.'s Mem.Supp. Sum.J. at 2-3). In other words, the Plaintiff claims that there is but one method of apportionment that is constitutionally acceptable — a system based on districts of equal population — and that all weighted voting plans are per se unconstitutional. In support of this position, the Plaintiff relies on Board of Estimate v. Morris, 489 U.S. 688, 109 S.Ct. 1433, 103 L.Ed.2d 717 (1989), and the recent interpretation of this case in Jackson v. Nassau County Bd. of Supervisors, 818 F. Supp. 509 (E.D.N.Y. 1993).
This Court finds Plaintiff's reliance on Morris to be misplaced, and that Plaintiff misinterprets the exhaustive analysis of Morris performed by Judge Spatt in Jackson. Furthermore, this Court finds that the current system of weighted voting in Delaware County satisfies the two core constitutional requirements for a plan of apportionment: (1) a quantitative requirement that guarantees that the vote of any citizen is approximately equal in weight to that of any other citizen; and (2) a qualitative requirement that the system provide fair and effective representation for all citizens. Morris, 489 U.S. at 701, 109 S.Ct. at 1442 (citing Reynolds v. Sims, 377 U.S. at 565-66, 576, 84 S.Ct. at 1383, 1389).
A. Weighted Voting Before Board of Estimate v. Morris
It is well settled that prior to Morris, both the Second Circuit and the Supreme Court had held that weighted voting plans substantially similar to that of Delaware County were constitutional. Abate v. Mundt, 403 U.S. 182, 91 S.Ct. 1904, 29 L.Ed.2d 399 (1971); League of Women Voters v. Nassau County Bd. of Supervisors, 737 F.2d 155 (2d Cir. 1984); Greenwald v. Board of Supervisors of the County of Sullivan, 567 F. Supp. 200 (S.D.N.Y. 1983), aff'd, 742 F.2d 1433 (2d Cir. 1983); Franklin v. Krause, 32 N.Y.2d 234, 344 N.Y.S.2d 885, 298 N.E.2d 68 (1973), appeal dismissed for want of substantial federal question, 415 U.S. 904, 94 S.Ct. 1397, 39 L.Ed.2d 461 (1974). Because these cases articulate the judicial approach to weighted voting and local government apportionment, and also because they are heavily drawn upon in the recent decisions in Morris and Jackson, it will be helpful to discus some of these cases in detail.
1. Weighted Voting based on the Banzhaf Index Franklin v. Krause was the leading weighted voting case before Morris. That case dealt with the plan of apportionment among the five towns of Nassau County, New York. Nassau County, like many of the New York counties, had rejected a weighted voting system based on a direct arithmetic relationship between weighted votes and population. This was because Hempstead, the most populous town in the county, had 56% of the county's total population and could therefore have dominated the Board of Supervisors whenever a simple majority of votes was required. In other words, the supervisor from Hempstead would have had 56% of the votes, but would have enjoyed 100% of the voting power in the great majority of board decisions.
In order to remedy the overvaluation of Hempstead's voting power and to assure that sparsely populated areas retained a voice in the democratic process, Nassau County employed a weighted voting plan based on a system known as the Banzhaf Index. This system used a computer driven statistical analysis to assign "voting power" to each representative, so that the power of each representative to affect the outcome of votes by casting the tie-breaking ballot in a theoretical number of possible voting combinations was equal to the percentage of the total county population of his or her town. For example, a representative whose district constitutes 15% of the county population would be assigned sufficient weighted votes to cast the tie-breaking vote in 15% of the board's decisions.
This approach was first recommended in Iannucci v. Board of Supervisors, 20 N.Y.2d 244, 282 N.Y.S.2d 502, 229 N.E.2d 195 (1967). In making this recommendation, the New York Court of Appeals held that other methods of apportioning weighted votes were not per se unconstitutional. However, the court cautioned that weighted voting plans might create their own power inequalities. The court gave as an example a situation in which 60% of the population elected one legislator, who would have effectively 100% of the power.
For a more detailed explanation of the Banzhaf Index, see Gefland Allbritton, supra note 3, at 112; Jackson v. Nassau County Bd. of Supervisors, 818 F. Supp. 509, 521-22 (E.D.N.Y. 1993) (quoting Justice White's explanation of the Banzhaf method from Whitcomb v. Chavis, 403 U.S. 124, 162-63, 91 S.Ct. 1858, 1859, 29 L.Ed.2d 363 (1971)); and Greenwald v. Board of Supervisors of the County of Sullivan, 567 F. Supp. 200, 203 (S.D.N.Y. 1983).
This system was declared unconstitutional, however, by the New York State Supreme Court. Franklin v. Krause, 72 Misc.2d 104, 338 N.Y.S.2d 561 (Sup.Ct. Nassau County 1972). The court criticized the "mathematical gyrations" used by the plan on the grounds that, while they preserved some voting power for the smaller political units, they did so at the expense of "the destruction of principles of majoritarean democracy." Id. at 108, 338 N.Y.S.2d at 565-66. The court concluded that, while this principle was not inviolate, the debasement of the rights of voters was not outweighed by the county's interest in the existing plan, as there were a number of viable alternative methods of reapportionment available to the county. Id.
The New York Court of Appeals disagreed and upheld the Nassau County weighted voting system. Franklin v. Krause, 32 N Y2d 234, 344 N.Y.S.2d 885, 298 N.E.2d 68 (1973). The Court began by recognizing that "the smaller communities are superenfranchised to a somewhat greater extent than the larger communities are disenfranchised." Id. at 237, 344 N.Y.S.2d 885, 298 N.E.2d 68. However, the court found that the range of deviation between each town's theoretical voting power and percentage of county population in each town was only 7.3%, and this fell within a constitutionally acceptable range based on prior case law. Id. In addition, the court found that the plan's requirement of "super-majorities," while admittedly being "artificial," prevented the constitutional infirmity of having one legislator with just over 50% of the votes and 100% of the power. Id.
The Court of Appeals methodology in calculating the range of deviation is important as it has been called into question by Morris.
The Court of Appeals first noted that the plan provided for a total of 130 votes to be distributed among the six supervisors, and then compared each town's percentage of voting power (as calculated by the theoretical Banzhaf Index) with each town's percentage of the total county population. For example, the court found that North Hempstead was the most underrepresented town in the county, as it constituted 16.5% of the county population, but had only 13% of the voting power, resulting in a deviation of -3.5%. At the other extreme was Glen Cove, with 1.8% of the county population and 5.6% of the voting power, producing a deviation of 3.8%. In order to calculate the range of deviation, the court simply calculated the difference between North Hempstead's -3.5% and Glen Cove's 3.8%, which produced a deviation range of 7.3%.
The Court of Appeals justified its decision to uphold the plan despite the deviation from the ideal population equality, on the grounds that "the small board, composed of the unit Supervisors, is the most efficient form of government, and has proved to be such over the years." Id. at 238, 344 N.Y.S.2d 885, 298 N.E.2d 68. The Court also recognized that the county had a legitimate interest in preserving traditional boundary lines, which facilitated local taxing and the delivery of local services. Id. at 241, 344 N.Y.S.2d 885, 298 N.E.2d 68. The Court of Appeals quoted the Supreme Court in Abate v. Mundt, 403 U.S. 182, 91 S.Ct. 1904, 29 L.Ed.2d 399 (1971), which upheld an apportionment plan with a deviation of 11.9%, and which recognized that local government (as opposed to state or congressional government) needed considerable flexibility in order to meet local needs, and that the "needs of a local community as a whole may sometimes justify departures from strict equality." Franklin, 32 N.Y.2d at 239, 344 N.Y.S.2d 885, 298 N.E.2d 68. The Court of Appeals concluded by reaffirming that the goal of apportionment plans was to remain one person, one vote, but added that the law had "assumed a desirable practicality because it allow[ed] for flexibility — something which, at least prior to Abate v. Mundt was lacking." Id. at 242, 344 N.Y.S.2d 885, 298 N.E.2d 68 (citation omitted).
The Supreme Court reviewed the case under its mandatory appellate jurisdiction, 28 U.S.C. § 1257(2), and dismissed it for want of a substantial federal question. Franklin v. Krause, 415 U.S. 904, 94 S.Ct. 1397, 39 L.Ed.2d 461 (1974). This type of dismissal is deemed to reach the merits of the case, and creates binding precedent. See League of Women Voters v. Nassau County Bd. of Supervisors, 737 F.2d 155 (2d Cir. 1984) (citing Fusari v. Steinberg, 419 U.S. 379, 391-92, 95 S.Ct. 533, 541, 42 L.Ed.2d 521 (1975) (Berger, C.J. concurring)).
2. Population-Based Apportionment Plans: The Multiple Representative Approach of Abate v. Mundt.
The reach and content of the precedent created by the Supreme Court's dismissal in Franklin was explored by the Second Circuit in League of Women Voters of Nassau County v. Nassau County Board of Supervisors, 737 F.2d 155 (2d Cir. 1983). This case was a subsequent challenge to the weighted voting system in Nassau County.
In League of Women Voters, the Second Circuit examined the questions presented to the Supreme Court in Franklin and found that the Court had created controlling precedent that weighted voting was not per se unconstitutional. Id. at 167. The Second Circuit concluded that it was bound by this precedent until such time as the Supreme Court informed them otherwise, either by a direct holding or by doctrinal developments. Id. at 171. In addition, the court reviewed a number of cases to show that "the theme" of the Supreme Court's jurisprudence in the area of local government legislative plans "has been flexibility" in allowing local governments to organize themselves to best meet the needs of the local communities. Id.
An alternative to the theoretical Banzhaf method of satisfying the one person, one vote command of Reynolds was affirmed by the Supreme Court in Abate v. Mundt, 403 U.S. 182, 91 S.Ct. 1904, 29 L.Ed.2d 399 (1971). In Abate the challenged plan maintained traditional political units of differing size in Rockland County, New York by assigning differing numbers of representatives to each district according to the district's population in relation to the population of the smallest town in the county. For example, the smallest town, Stony Point, had a population of 12,114 and was assigned one representative. Orangetown, with a population of 52,080 was 4.3 times larger than Stony Point and was therefore assigned 4 representatives (fractional results being rounded to the nearest integer). The deviation from ideal population under this plan was 11.9%. Again, the Supreme Court supported its decision to affirm the plan by citing the need for flexibility and a desire to preserve traditional political units. Id. at 185, 91 S.Ct. at 1906 ("viable local governments may need considerable flexibility in municipal arrangements if they are to meet changing societal needs, . . . and that a desire to preserve the integrity of political subdivisions may justify an apportionment plan which departs from numerical equality").
The deviation was the result of the rounding process, and was calculated by taking the total population of the county (218,804) and dividing it by the number of representatives (18) to arrive at the mean or ideal population (12,155). The ideal population was then multiplied by the number of representatives assigned to the town, and the product was subtracted from each town's population. The result was then divided by the adjusted mean/ideal population to arrive at the percentage of deviation from ideal. For example, Orangetown had a population of 52,080, and had 4 representatives. Subtracting the adjusted mean/ideal population of 48,620 (4 x 12,155) produced 3460. By dividing this by the adjusted mean/ideal population (3460 ÷ 48,620) the Court produced a deviation of -7.1%. To calculate the maximum deviation, the Court calculated the difference between the most underrepresented and most over represented towns (-7.1 Orangetown and 4.8 Clarkstown = 11.9%).
The desire to provide for the efficient delivery of local services and the maintenance of traditional political units that formed the purpose behind the multiple representative system in Abate also underlies weighted voting systems. Both systems are designed to satisfy the quantitative command of one person, one vote by giving all citizens a vote of equivalent weight, while preserving the traditional political units. The two approaches differ in the method used to achieve equality of voting weight; in Abate the plan varies the number of representatives in each district in proportion to the size of the population, whereas in weighted voting, the number of representatives is held constant, and the votes of each representative are increased in proportion to the population of their district. The latter is designed to maintain a smaller and more efficient legislature in counties in which the difference in population between the largest and smallest towns would require the election of a large number of representatives in order to ensure that the smallest town was assigned one representative. In the instant case for example, if Bovina with a population of 550 was assigned one representative, Sidney, with a population of 6667, would require twelve representatives. The legislature would be substantially larger than it is today, and the efficiency would likely suffer.
Weighted voting plans may be distinguished by the method used to apportion the weighted votes among representatives. In the simple arithmetic model, the number of actual votes are distributed in proportion to the percentage of the total district population in each unit. In a Banzhaf-based model, the weighted votes are apportioned to each representative based on the theoretical voting power each representative should possess.
The law prior to Board of Estimate v. Morris was therefore clearly contrary to the Plaintiff's position. The Supreme Court had rejected the contention that weighted voting was per se unconstitutional in Franklin v. Krause. Jurisprudence with regard to local government apportionment systems was characterized by judicial flexibility and concern with maintaining efficient local government and preserving traditional political units. The nature of the inquiry when investigating the constitutionality of local apportionment plans was based upon the one person, one vote principle as expressed in Reynolds, which stated that the Constitution required that the vote of any citizen be approximately equal in weight to that of any other citizen. The aim being to provide fair and effective representation for all citizens. Reynolds, 377 U.S. at 579, 84 S.Ct. at 1390.
Fair and effective representation has been described by Professor Tribe in the following way: "There is a guarantee of some form of mathematical equality: every individual has the right to have her district represented in proportion to its population. There is, as well, a more elusive guarantee of fair representation: certain mathematically palatable apportionment schemes will be overturned because they systematically circumscribe the voting impact of specific population groups." L. Tribe, American Constitutional Law 1063 (2d ed. 1988) (quoted in Gefland Allbritton, supra note 3, at 119 n. 137).
In order to bring this analysis of the law up-to-date and applying it to the case at hand, the Court must answer two questions: (1) what did Board of Estimate v. Morris say about weighted voting; and (2) how does the Delaware County weighted voting system comport with the constitutional standard?
B. Board of Estimate v. Morris
The answer to the first question, what did Morris say about weighted voting, is very difficult to divine. The fact is that Morris was not a true weighted voting case, and therefore the Court did not face head-on the issue of the constitutionality of weighted voting. The Court, however, did have occasion to address the use of the Banzhaf Index and clearly rejected it, at least as a mechanism for measuring the deviation of voting districts from population equality, and by extension, also as a method of structuring a weighted voting plan. In order to understand exactly what the Supreme Court's decision in Morris meant for weighted voting, it is necessary to examine the structure of the New York Board of Estimate and see how the defendants sought to use the Banzhaf Index as a defense.
The Board of Estimate consisted of the five New York borough presidents, plus three officials elected citywide — the mayor, the comptroller, and the city council president. Each of the borough presidents had a single vote, even though their electoral districts ranged in size from Brooklyn, population 2,230,936, to Staten Island, population 352,151. The three citywide members were entitled to cast two votes each, except the mayor was not entitled to vote on budgetary matters. Morris, 489 U.S. at 690-91, 109 S.Ct. at 1436.
The difference in the size of the boroughs combined with the fact that each borough elected one representative who was entitled to cast a single vote, produced a maximum deviation of 132.9% from the ideal population. Even when the citywide members were included in the calculation, the deviation between votes actually assigned to each district and votes that should have been assigned under an ideal system with population equality remained high at 78%. Id. at 701, 109 S.Ct. at 1442. The Morris Court noted that "no case of ours has indicated that a deviation of some 78% could ever be justified. . . . At the very least, the local government seeking to support such a difference between electoral districts would bear a very difficult burden, and we are not prepared to differ with the holding of the courts below that this burden has not been carried." Id. at 702, 109 S.Ct. at 1442. The Court made it clear that the board's attempt to justify the existing system based on "exigencies of history and convenience" could not outweigh the debasement of the voters' constitutional right of one person, one vote. Id. at 703 n. 10, 109 S.Ct. at 1443 n. 10.
The Supreme Court simply calculated the mean borough population (1,414,206) and then determined the percentage by which the most populous district, Brooklyn, population 2,230,936, exceeded the mean, and the percentage by which the least populous borough, Staten Island, fell below the mean.
Faced with the violation of the one person, one vote principle of Reynolds, the board first attempted to defend its system on the grounds that the board was a nonelective, nonlegislative body, and therefore was not subject to the one person, one vote rule. However both the New York Court of Appeals and the Supreme Court rejected that argument and found that the board was a legislative body, and was therefore subject to the Fourteenth Amendment. Morris, 489 U.S. at 692, 109 S.Ct. at 1437.
The board then attempted to argue that the courts below had used the incorrect methodology in calculating the deviation from the ideal population. The board argued that the citywide members should be included and the Banzhaf Index should be used to calculate the deviation from the ideal system. When the Banzhaf Index was used to calculate the power of each representative to cast a tie-breaking vote on the board, the deviation from ideal population decreased to 30.8% on nonbudget matters and, because the mayor could not vote on budgetary matters, higher on budget issues. Id. at 697-98, 109 S.Ct. at 1440.
Both the Court of Appeals and the Supreme Court rejected the board's use of the Banzhaf method in this context. They noted that the Banzhaf method was purely theoretical and failed to consider how the board actually worked in practice. The Supreme Court also criticized the method on the grounds that it was unrealistic in that it did not consider political affiliation, race or any other factors which affected the actual voting power of the electorate. Id. at 698-99, 109 S.Ct. at 1440-41. The Supreme Court explained:
The personal right to vote is a value in itself, and a citizen is, without more and without mathematically calculating his power to determine the outcome of an election, shortchanged if he may vote for only one representative when citizens in a neighboring district, of equal population, vote for two; or to put it another way, if he may vote for one representative and the voters in another district half the size also elect one representative.Id. at 698, 109 S.Ct. at 1440.
Justice White did not take the opportunity to consider whether this constitutional infirmity could be cured by having the citizens of the larger district elect two representatives to every one representative elected by the district half its size, or by having the representative elected from the smaller of the two districts assigned a vote in the legislature worth half of that of the representative from the district twice its size. Justice White noted that the Court was not presented with the question of the constitutionality of alternative board structures. Morris, 489 U.S. at 703 n. 11, 109 S.Ct. at 1443 n. 11.
When the Court rejected the Banzhaf Index as a measure of deviation from ideal population, the Court reaffirmed the "population-based approach of . . . cases from Reynolds through Abate." Id. at 698, 109 S.Ct. at 1440. The Court then went on to explain exactly how these cases should be approached: "In calculating the deviation among districts, the relevant inquiry is whether `the vote of any citizen is approximately equal in weight to that of any other citizen,'. . . the aim being to provide `fair and effective representation for all citizens.'" Id. at 701, 109 S.Ct. at 1442 (citations omitted) (quoting Reynolds, 377 U.S. at 565-66, 84 S.Ct. at 1383).
The emphasis in Morris on using population as the basis of an apportionment plan, and its concern with ensuring relative equality in the weight of each citizen's vote, would clearly support weighted voting systems which adopt a population-based approach. That is, those plans that allocate votes to representatives in proportion to the population of their districts, and thereby ensuring that the weight of each citizen's vote is approximately equivalent to that of all other citizens. This approach can be contrasted with those weighted voting plans that use the theoretical Banzhaf Index to allocate votes based on the relative voting power of each district. The language in Morris clearly rejected the theoretical validity of the Banzhaf Index. This rejection not only discards the Banzhaf Index as a means of calculating the deviation of a plan from population equality, but also questions the continued viability of those weighted voting plans that are based on a theoretical voting power model.
The language in Morris also expresses the Court's concern with the qualitative goal of assuring fair and effective representation for all citizens. Thus, the Court made it clear that even if a plan is designed on a population-based model, the plan will only be constitutional as long as (1) its deviation from population equality is within a constitutionally acceptable degree, and (2) it does not substantially circumscribe or overvalue the votes of any particular population groups by, for example, granting a representative with 51% of the population 100% of the power, thereby denying the smaller political units an effective voice in government.
In reading Morris, there is no reason to adopt the Plaintiff's view that the Supreme Court has turned away from a flexible approach to local government organization, and is now sanctioning apportionment based on districts with equal populations as the only constitutionally acceptable system. While Morris includes language that may indicate a preference for a system based on equal population districts, the Court chose to focus the relevant inquiry on the comparative "weight" of the citizens' votes. Had the Court intended to rely exclusively on districts of equal population, it would surely have based the relevant inquiry simply on the population of each district. In addition, the Morris Court reaffirmed the population-based approaches to apportionment in Reynolds and Abate. The Rockland County apportionment scheme that was upheld in Abate v. Mundt was based on districts with unequal populations. This system is theoretically similar to weighted voting, and serves the same purposes — providing all citizens a vote of equal weight, while maintaining traditional political units. Thus, while the Supreme Court may prefer electoral districts of equal population, local governments still have the flexibility to organize themselves in ways that meet the needs of the local communities, as long as they are within reasonable constitutional parameters.
There is also nothing in the language of Morris to support the Plaintiff's contention that the board's interest in maintaining traditional political units can no longer justify deviation from population equality (Pl.'s Mem.Supp.Sum.J. at 3). While it is true that in Morris the Court rejected the board's argument that history and efficiency justified a deviation from population equality of 78%, it does not necessarily follow that a desire to maintain traditional political units and local government efficiency can no longer justify smaller deviations from pure equality as they did in Abate and Franklin. In Abate the Court made it clear that while tradition, efficiency and absence of built-in bias could justify the 11.9% deviation from equality found in the plan, the Court added that "nothing we say today should be taken to imply that even these factors could justify a substantially greater deviation from population equality." Id. at 187, 91 S.Ct. at 1908. Thus, in rejecting these factors when used to justify a deviation of 78%, the Court in Morris made no doctrinal change in its approach to these factors.
The Plaintiff relies heavily on language in Jackson v. Nassau County Board of Supervisors, 818 F. Supp. 509 (E.D.N.Y. 1993), to find support for the proposition that the Supreme Court in Morris made all types of weighted voting per se unconstitutional (Pl.'s Mem.Supp.Sum.J. at 1-3). While the decision in Jackson is, of course, merely persuasive authority in this Court, Judge Spatt's exhaustive analysis of the history of weighted voting and his interpretation of the Supreme Court's decision in Morris is particularly authoritative and deserves careful consideration. In support of the contention that all weighted voting plans are unconstitutional, the Plaintiff quotes the paragraph in Jackson in which Judge Spatt states:
the Supreme Court firmly rejected weighted voting, not only because of the mathematical quagmire such a system engenders, but just as importantly because the methodology fails to take into account other critical factors related to the actual daily operations of a governing body. There is no question that the Supreme Court took the opportunity to express not only a preference, but a directive that legislators be elected by and represent citizens in districts of substantially equal size.Id. at 532.
However, it is clear from the context of the quoted passage that Judge Spatt was referring specifically to weighted voting systems based on a theoretical Banzhaf Index, and not necessarily to other types of weighted voting plans. This is apparent from references in the quotation itself to the "mathematical quagmire" and the failure of weighted voting to consider the actual, as opposed to the theoretical daily operations of the legislature. Judge Spatt's focus is understandable, as he was dealing with a weighted voting system based on a Banzhaf Index in Nassau County.
While Judge Spatt uses sweeping language to condemn weighted voting plans based on the Banzhaf model, he does not go so far as to state that all weighted voting plans are per se unconstitutional. In fact, Judge Spatt distinguishes his position in Jackson from scholars who believe that weighted voting was made per se unconstitutional in Morris. In discussing the opinion of Professor Scarrow, a political science professor at the State University of New York at Stony Brook, Judge Spatt opined that "[a]ccording to Prof. Scarrow, in the wake of Morris weighted voting can no longer be used. This Court agrees, at least in so far as the `mathematical quagmire' presently existing in Nassau County is concerned." Id. at 535 (emphasis added). Judge Spatt then described the basic flaws in the structure of the Nassau County system: (1) the system was artificial and contrived, (2) it was incomprehensible to most citizens, (3) it could not work with a simple majority, which is a fundamental part of most democratic systems, and (4) the "supermajorities" insured that Hempstead voters were permanently underrepresented in relation to their population. Based on these flaws, Judge Spatt held that "weighted voting as presently utilized by the Nassau County Board of Supervisors is unconstitutional." Id. at 535 (emphasis added). Thus, Judge Spatt's holding was narrowly tailored to the facts of the weighted voting system in Nassau County and he did not simply declare the Nassau County system per se unconstitutional based on Morris.
What then is the state of the law after Board of Estimate v. Morris? Weighted voting systems, like that in Nassau County, which are based on a theoretical voting power analysis and which do not provide fair and effective representation for all citizens have been rejected. By extension, the method of calculating the deviation from population equality performed by the Court of Appeals in Franklin v. Krause, which compared each representative's theoretical voting power with the population of his or her district, has also been rejected. Replacing it is a system based on the Reynolds-Abate line of cases, which relied instead on a simpler population-based approach.
Nothing in the Morris decision indicates a departure from the Court's emphasis on flexibility in local government organization, or a departure from the Court's concern with maintaining local government efficiency and traditional political units. The Court affirmed the standard articulated in Reynolds, that the relevant inquiry when testing the constitutionality of apportionment plans is that the vote of any citizen must be approximately equal in weight to that of any other citizen, and that the goal is to provide fair and effective representation to all citizens. The final question before the Court remains: does the weighted voting system in Delaware County conform to current constitutional requirements?
C. The Constitutionality of the Delaware County Weighted Voting System
In approaching the question of the constitutionality of a local government apportionment plan, such as the Delaware County plan at issue here, it is useful to split the inquiry into two parts: first, a quantitative inquiry to determine whether the weight of each citizen's vote is approximately equal to that of every other citizen; and second, a qualitative inquiry to ensure that all citizens are provided fair and effective representation. As for the quantitative inquiry, if the analysis reveals a degree of deviation from population equality, this part of the inquiry should also test to see if the degree of deviation falls within the constitutionally acceptable range. This requires the court to determine whether the degree of deviation from population equality can be justified by legitimate government interests. As for the qualitative inquiry, factors to be considered in this analysis include evidence of built-in bias against any particular political interest or group; whether the method of apportionment is so complex as to confuse and alienate the voters; whether one representative has greater than 50% of the votes and effectively 100% of the power, or alternatively, whether the smaller political units are denied an effective voice in board decisions.
1. Quantitative Analysis
There is no question in this case that the town supervisors are not elected by districts of substantially similar size. The Plaintiff's analysis, which is methodologically based on the approach of the Supreme Court in Abate and the Eastern District of New York in Jackson, reveals that the population of the largest town, Sidney, is 168.2% above the mean population, while that of the smallest town, Bovina, is 77.9% below the mean. Using a strict population-based analysis, this approach produces a total deviation of 246.1%.
This Court, however, finds that this method of analysis is not the relevant measure of whether the votes of the citizens of Delaware County are weighted equally in this case. Plaintiff's method of calculating the deviation of an apportionment plan was developed in Abate. In Abate the calculation was used to test a system in which the voting weight of citizens in districts of unequal size was equalized by assigning additional representatives to the larger districts based on the ratio of the town's population to the population of the smallest district, which was assigned one representative. In other words, this method of apportionment was used to satisfy the requirements of one person, one vote by weighting the votes of each district in proportion to its population by assigning more representatives with a single vote to each district, rather than keeping the number of representatives constant and weighting each representative's number of votes in proportion to the population of the district. If Delaware County was required to use the Abate method to achieve its goal of retaining the traditional political units, the Board of Supervisors would be large and unwieldy, with over 85 members. To force Delaware County into this inefficient method of apportionment would run counter to the very flexibility that the Supreme Court was trying to encourage in Abate. Inflexibly using this method to calculate the deviation from population equality of all electoral systems will simply invalidate all systems other than an Abate-type multiple representative plan or a plan designed around equal population districts. Because this approach frustrates the objective of encouraging flexibility and efficiency in local government organization, this Court declines to adopt the position that a single test is appropriate in all situations.
Being mindful of the need for flexibility and taking into consideration the local circumstances, the Court finds that the analysis employed by the Defendant is the more appropriate test in this case. The analysis employed by the board is a modification of the test used by the Court of Appeals in Franklin. In this case the board compares the percentage of actual votes allocated to each representative against the percentage of the county population in each district, whereas in Franklin the court compared the theoretical voting power of each representative against the population in each district. Using the Defendant's analysis reveals that the most over represented town is Andes, with 2.737% of the total population and 2.988% of the votes, producing a deviation of 0.25. While the most underrepresented town is Sidney, with 14.117% of the total population and 13.448% of the total votes, producing a deviation of -0.67. Using this method, the total deviation is less than 1% (0.92) from population equality.
This important distinction appears to have been missed by the Defendant, which states that its test is based on the Krause analysis, and continues to use the loaded term "voting power" when referring to a simple arithmetic analysis of the distribution of votes (Def.'s Brief, Point II).
This deviation is clearly within the permissible constitutional range, as defined by Abate and its progeny. This point is effectively conceded by the Plaintiff in the Statement Pursuant to Local Rule 10(j), in which she states that "[w]hen voting, the board uses a weighting system under which each member casts a vote which is essentially proportionate to the population of his or her town." (Pl.'s Statement Pursuant to Local Rule 10(j), at ¶ 9).
In addition, the Defendant offers the following arguments as justification for the weighted voting plan: (1) that the natural and historical political units in the county have been the nineteen towns that today form the electoral districts; (2) that citizens identify with their towns through the school PTAs and other civic organizations; and (3) that this identification of the voters with their political units would be lost if a system of equal population districts split some communities and artificially created others. Based on the very small deviation from population equality, and the board's interest in maintaining the traditional political units, this Court finds that the current system of weighted voting satisfies the quantitative one person, one vote requirement for apportionment plans.
2. Qualitative Analysis
The weighted voting system in Delaware County appears to suffer from none of the qualitative constitutional infirmities that plagued Nassau County. The system of apportioning votes to town supervisors in proportion to the town population is relatively easy to understand, and has none of the "mathematical gyrations" that made the Nassau County system incomprehensible to all but the experts. It is based on real votes and real populations, rather than on a theoretical model of voting coalitions. Most importantly, no single town has close to 50% of the votes, and no supervisor can therefore dominate the board. As a result, the weighted voting plan relies on the fundamental democratic practice of having simple majorities decide the great majority of its votes, and conventionally calculated "supermajorities" decide on a small number of particularly important measures. No single group of voters are permanently underrepresented in proportion to their numbers, and even the smallest political unit has a meaningful say in the decisions of the board. In addition, the Plaintiff has not suggested that the plan contains any built-in bias favoring a particular political interest, race or geographic area.
In short, the details of the plan and the distribution of the votes ensure that there is nothing to prevent each citizen from receiving fair and effective representation under the current weighted voting plan. This conclusion is supported by the fact that the Plaintiff's papers have made no allegations of bias or legislative inefficiency caused by the plan.
III.
Weighted voting plans which apportion votes in proportion to the populations of the electoral districts were not made per se unconstitutional by the Supreme Court in Board of Estimate v. Morris. The constitutionality of each plan can only be determined by an individualized analysis of the challenged plan. This inquiry must include a quantitative one person, one vote analysis to ensure that the vote of each citizen is approximately equal in weight to that of any other citizen, and a qualitative analysis to ensure that each citizen receives fair and effective representation. The court conducting this inquiry should be guided by the desirability of affording local governments considerable flexibility in structuring municipal arrangements to meet the changing needs of their communities, and to preserve traditional political units. Because the Delaware County weighted voting system satisfies each of these requirements, this Court holds that the plan is constitutional.
The papers submitted by all parties reveal that there are no issues of material fact in dispute in this case. There was merely the one question of law regarding whether the system of weighted voting in Delaware County was constitutional under the Fourteenth Amendment's Equal Protection Clause. In situations in which there is no issue of material fact in dispute and the question of law has been resolved by the court, "the weight of authority is that summary judgment may be rendered in favor of the opposing party even though that party has made no cross-motion under Rule 56." Charles A. Wright et al., 10A Federal Practice and Procedure § 2720, at 29 (2d ed. 1983). Accordingly, this Court finds that there are no material issues of fact in dispute, and that the Plaintiff's motion for summary judgment is rejected and the Defendant, the Delaware County Board of Supervisors, is entitled to summary judgment as a matter of law.
IT IS SO ORDERED.
Appendix 1
1990 TABLE A TABLE B TABLE C TOWN POP. (Simple) (2/3s) (3/5s)
Andes 1291 104 65 62 Bovina 550 39 28 27 Colchester 1928 146 98 93 Davenport 2438 182 124 117 Delhi 5015 364 257 236 Deposit 1824 142 92 88 Franklin 2471 183 126 119 Hamden 1144 78 58 56 Hancock 3384 253 172 161 Harpersfield 1450 109 74 70 Kortright 1410 108 72 68 Masonville 1352 106 69 65 Meredith 1513 111 77 73 Middletown 3406 254 174 162 Roxbury 2388 181 121 115 Sidney 6667 468 358 308 Stamford 2047 149 104 99 Tompkins 994 74 51 48 Walton 5953 429 308 277
Totals 47,225 3480 2428 2244
Majority 1741 1619 1347
Adopted 7/10/91; effective 8/24/91
Appendix 2
TOWN % OF POPULATION % OF VOTE % OF VOTE % OF VOTE TABLE A TABLE B TABLE C
Andes 2.7337 2.9885 2.6771 2.7629 Bovina 1.1646 1.1207 1.1532 1.2032 Colchester 4.0826 4.1954 4.0362 4.1444 Davenport 5.1625 5.2299 5.1071 5.2139 Delhi 10.6194fn_ 10.4598 10.5848 10.5169 Deposit 3.8624 4.0805 3.7891 3.9216 Franklin 5.2324 5.2586 5.1895 5.3030 Hamden 2.4224 2.2414 2.3888 2.4955 Hancock 7.1657 7.2701 7.0840 7.1747 Harpersfield 3.0704 3.1322 3.0478 3.1194 Kortright 2.9857 3.1034 2.9654 3.0303 Masonville 2.8629 3.0460 2.8418 2.8966 Meredith 3.2038 3.1897 3.1713 3.2531 Middletown 7.2123 7.2989 7.1664 7.2193 Roxbury 5.0566 5.2011 4.9835 5.1248 Sidney 14.1175 13.4483 14.7446 13.7255 Stamford 4.3346 4.2816 4.2834 4.4118 Tompkins 2.1048 2.1264 2.1005 2.1390 Walton 12.6056 12.3276 12.6853 12.3440
The court has recalculated these numbers.
Figures contained in Local Law #2 of 1991 expressed as percentage