Opinion
SJC-08715
March 7, 2002.
April 29, 2002.
Present: Marshall, C.J., Greaney, Ireland, Spina, Cowin, Sosman, Cordy, JJ.
Elections. Constitutional Law, Elections, Apportionment of representation, Equal protection of laws, Redistricting, Representative districts. Federal Voting Rights Act.
Discussion of the constitutional and statutory requirements involved in the division of the Commonwealth into 160 representative districts. [616-618]
Statement of the relevant characteristics of the 2001 redistricting plan for the 160 representatives districts of the House of Representatives in the Commonwealth, and the progression and evolution of its predecessor plans. [618-621]
Statement of an alternative plan to St. 2001, c. 125, § (the redistricting plan) presented by plaintiffs seeking to invalidate the 2001 redistricting plan for the 160 representative districts of the House of Representatives in the Commonwealth. [621-622]
Statutes 2001, c. 125, § 1 (the redistricitng statute), which established the representative districts of the House of Representatives in the Commonwealth, did not violate the directive of art. 101 of the Amendments to the Constitution of the Commonwealth, because it placed portions of the city of Chelmsford in four representative districts when all of the requirements of art. 101 could have been met with fewer divisions, where the Legislature reasonably could have concluded that it was preferable to seek a closer approximation of population equality in the districts in Chelmsford area at the expense of dividing the town among four districts, as part of a Statewide plan that sought to balance these conflicting objectives. [622-625] Sosman J., with Greaney and Cowin, JJ., joined, dissenting.
Plaintiffs seeking to invalidate St. 2001, c. 125, § 1 (the redistricing plan), which established the representative districts of the house of Representatives in the Commonwealth, failed to demonstrate that the Legislature's redistricting plan had a discriminatory effect so as to constitute a partisan political gerrymander in violation of the Fourteenth Amendment to the United States Constitution, or that the Legislature had made purposeful, concerted effort to disfavor a political party through the redistricting process. [625-626]
Civil action commenced in the Supreme Judicial Court for the Commonwealth on November 16, 2001.
After referral to the Supreme Judicial Court for the county of Suffolk, the action was returned to the Commonwealth court by Greaney, J., to proceed on an expedited basis.
Richard P. McClure, pro se.
Peter Sacks, Assistant Attorney General (Lawrence S. DiCara, Special Assistant Attorney General, with him) for Secretary of the Commonwealth.
This is another dispute arising out of legislative redistricting for the 2002 general election. See Mayor of Cambridge v. Secretary of the Commonwealth, ante 476, 477 (2002). In addition to declaratory and injunctive relief, the plaintiffs, registered voters living in various precincts of the town of Chelmsford, seek relief in the nature of mandamus against the defendant, Secretary of the Commonwealth, that would invalidate the 2001 redistricting plan for the 160 representative districts of the House of Representatives. The plaintiffs maintain that St. 2001, c. 125, § 1 (the redistricting statute), which establishes the new districts, is unconstitutional under art. 101, as amended by arts. 109, 117, and 119 of the Amendments to the Constitution of the Commonwealth, because it places portions of Chelmsford in four representative districts when art. 101's requirements could have been satisfied by dividing it into fewer districts. The plaintiffs also claim that the redistricting constitutes unconstitutional partisan gerrymandering in violation of the equal protection clause of the Fourteenth Amendment to the United States Constitution. We conclude that the plaintiffs have failed to demonstrate that the redistricting statute violates either the Massachusetts or the United States Constitution. Accordingly, the defendant is entitled to judgment.
Article 101 of the Amendments to the Constitution of the Commonwealth, as amended, states, in part, that the Legislature shall "divide the Commonwealth into one hundred and sixty representative districts of contiguous territory so that each representative will represent an equal number of inhabitants, as nearly as may be; and such districts shall be formed, as nearly as may be, without uniting two counties or parts of two or more counties, two towns or parts of two or more towns, two cities or parts of two or more cities, or a city and a town, or parts of cities and towns, into one district. Such districts shall also be so formed that no town containing less than twenty-five hundred inhabitants . . . shall be divided."
The plaintiffs also claim in a single sentence that the redistricting statute violates due process. Their claim does not rise to the level of acceptable appellate argument. See Adoption of Kimberly, 414 Mass. 526, 536-537 (1993).
1. Procedural background. Section 2 of the redistricting statute vests jurisdiction in the Supreme Judicial Court for "any petition for a writ of mandamus relative to the establishment of 160 representative districts under section 1 [of this act]." See also art. 101, §§ 1 and 3. Accordingly, the plaintiffs filed their complaint in this court. The case was referred to the county court to be prepared for disposition "as promptly as possible." The parties filed a statement of agreed facts. A single justice then ordered that the case proceed in this court on an expedited basis.
2. Constitutional and statutory requirements. Article 101, adopted in 1974, reduced the House of Representatives from 240 to 160 members, eliminated the last vestiges of territorial representation, and firmly established population equality as the main organizing principle for legislative districts. As we stated in Brookline v. Secretary of the Commonwealth, 417 Mass. 406, 423-424 (1994), "art. 101 directs that each representative `represent an equal number of inhabitants,' not cities and towns, thereby requiring equal representation, in the sense that persons are placed in a district that has the same number of inhabitants as every other district, `as nearly as may be.'" As subsequently amended, art. 101 also requires the Legislature to redivide the Commonwealth into new representative, senatorial, and councillor districts, in the year following each Federal census. The purpose for this mandatory decennial redivision is to ensure that districts are appropriately adjusted to meet population changes occurring in the ten-year period between each Federal census.
The Constitution formerly provided for cities and towns themselves to be represented in the House of Representatives. Part II, c. 1, § 3, art. 2 of the Massachusetts Constitution. Arts. 12, 13 of the Amendments to the Massachusetts Constitution. Municipal representation ended, however, in 1857, with the adoption of art. 21 of the Amendments that required that representation be apportioned to and within counties, "equally, as nearly as may be," based on the numbers of "legal voters." In 1970, art. 92 of the Amendments was ratified, requiring that every representative district in the Commonwealth comprise "an equal number of inhabitants, as nearly as may be," and ending apportionment by county except that Dukes and Nantucket counties were each made a separate district. In 1974, art. 101 eliminated those exceptions.
In dividing the Commonwealth into 160 representative districts, the Legislature must conform to State and Federal constitutional and statutory requirements, some of which are inherently contradictory. See Mayor of Cambridge v. Secretary of the Commonwealth,supra at 478-479. Article 101 mandates that the Legislature's division of the Commonwealth results in districts "[1] of contiguous territory [2] so that each representative will represent an equal number of inhabitants, as nearly as may be; [3] and such districts shall be formed, as nearly as may be, without uniting two counties or parts of two or more counties, two towns or parts of two or more towns, two cities or parts of two or more cities, or a city and a town, or parts of cities and towns, into one district[; and 4 such] districts shall also be formed that no town containing less than twenty-five hundred inhabitants . . . shall be divided." Merriam v. Secretary of the Commonwealth, 375 Mass. 246, 248 (1978). See Brookline v. Secretary of the Commonwealth, 417 Mass. 406, 413-414 (1994).
The equal protection provision of the Fourteenth Amendment independently imposes an equal representation requirement on electoral districting. Reynolds v. Sims, 377 U.S. 533, 577 (1964). The United States Supreme Court has stated that an apportionment plan with a maximum population deviation of under 10% is a deviation that the Legislature need not justify. Voinovich v. Quilter, 507 U.S. 146, 161 (1993), quotingBrown v. Thomson, 462 U.S. 835, 842-843 (1983). See Brookline v.Secretary of the Commonwealth, supra at 414, and cases cited. The maximum population deviation is calculated by determining the range of population deviation between the largest and smallest districts from the "ideal population" of a district. Mayor of Cambridge v. Secretary of the Commonwealth, supra at 478. See also Black Political Task Force v.Connolly, 679 F. Supp. 109, 114 (D.Mass. 1988). Thus, where a plan includes no district with a population more than 5% under or 5% over the "ideal district population," the plan is within the 10% range and thus meets Federal population equality requirements (± 5% standard). The Federal Voting Rights Act of 1965 must also be considered by the Legislature when implementing a redistricting plan. Brookline v.Secretary of the Commonwealth, supra at 414-415. See Mayor of Cambridge v. Secretary of the Commonwealth, supra at 479.
The "ideal population" of a district is the mathematical division of the number of districts into the total population.
A plan, of course, can still meet Federal population requirements even if it had a district that was 6% over the ideal population so long as no other district was more than 4%.
3. The 2001 redistricting plan. The 2001 redistricting statute and the legislative deliberation that led to its enactment cannot be adequately understood or evaluated without examining the progression and evolution of its predecessors. Between the adoption of art. 101 in 1974 and the enactment of the redistricting statute, the Legislature passed four successive redistricting plans. Each successive plan contained more districts that united parts of or whole cities, towns, or counties than the preceding plan, and each moved closer to population equality between districts, with only the last plan meeting the Federal ± 5% standard.
The first plan (St. 1977, c. 277) was based on the 1975 State census. It included ninety-two (out of 160) districts that did not meet the art. 101 goal of avoiding combinations of political subdivisions. Even with all of these combinations the plan included districts that deviated from the ideal population by more than ± 5%. Brookline v.Secretary of the Commonwealth, supra at 415-416 n. 10.
The use of the State census was abolished by art. 117 of the Amendments to the Massachusetts Constitution, which amended art. 101 in 1990 to require the Federal census be used as the basis for determining districts.
The second plan (St. 1987, c. 341, as amended by St. 1987, c. 715) was based on the 1985 State census. It included ninety-five districts that did not meet the art. 101 goal of avoiding combinations of political subdivisions. But, like the 1977 plan, it also included districts that deviated from the ideal population by more than ± 5%. Id. The plan was challenged in Federal court on equal protection ("one person one vote") grounds in Black Political Task Force v. Connolly, supra. The population deviations between districts (including a maximum deviation of 21.9%, id. at 124) were unsuccessfully justified in that case by the Secretary of the Commonwealth, who relied on art. 101's instruction to minimize the creation of districts that combined political subdivisions. The court concluded that art. 101's stated policy of minimizing the division and unification of cities and towns did not provide sufficient justification for the deviations from population equality and, accordingly, struck down the 1987 plan. Id. at 130-131.
In response to the decision in the Black Political Task Force case, the Legislature enacted a third plan (St. 1988, c. 11) that included one hundred districts that did not meet the art. 101 goal of avoiding combinations of political subdivisions. Yet the plan still contained some districts that deviated from the ideal population by the ± 5% standard. Brookline v. Secretary of the Commonwealth, supra. Only two years later, when the 1990 Federal census was completed, ninety of the 160 districts were of questionable validity under Federal population equality standards. Id. at 411.
The fourth plan (St. 1993, c. 273, § 1) included 114 districts that did not meet the art. 101 goal of avoiding combinations of political subdivisions. However, unlike prior plans, this plan did not contain any districts that deviated from the ideal population by more than ± 5%. Id. at 415. It was challenged by the town of Brookline for excessively dividing certain municipalities to further the goal of population equality and minority representation. In denying the challenge we stated:
"We reject the contention that, in enacting [the plan], the Legislature erred by giving undue emphasis to population equality among districts. . . . The principle of electoral equality has long been recognized as one of the fundamental rights protected by our Constitution. In 1916, Chief Justice Rugg, speaking of art. 21 of the Amendments to our Constitution, observed: `The great principle established by this amendment is equality of representation among all the [citizens] of the Commonwealth. That is a fundamental principle of representative government. . . . There can be no equality among citizens if the vote of one counts for considerably more than that of another in electing public officials.' . . . In art. 101, the people placed population equality on at least the same footing with the respect expressed for the cohesiveness of political subdivisions."
Id. at 418, quoting Attorney Gen. v. Suffolk County Apportionment Comm'rs, 224 Mass. 598, 604 (1916).
The 2000 Federal census results revealed that, as a result of population growth and shifts, the representative districts established in 1993 once again no longer met Federal population equality standards and had to be significantly altered. The census showed that Commonwealth's population had grown to 6,349,097 inhabitants, yielding a new ideal representative district population of 39,682 persons. This meant that, for Federal constitutional purposes, the presumptively permissible range for a district was now 39,682 ± 5%, i.e., between 37,698 and 41,666 persons. The ideal population under the 1993 plan had been 37,603 persons, which now fell outside of the presumptively permissible range. Due to uneven population growth in different communities, the districts established in 1993 now deviated from the new ideal population by from -15.24% to +28.33% for a maximum deviation of 43.57%.
Based on the 2000 census results, nearly 200 cities and towns also redrew their ward or precinct lines. These new "building blocks" became the basis for the new districts to be drawn by the Legislature. They vary in size from many hundred to many thousand, are therefore not interchangeable in size and substantially complicate the task of building districts that are both compact and equal in population.
The Legislature established a Joint Special Committee to recommend redistricting plans for House, Senate, Executive Council, and Congressional districts. After holding public hearings, the Committee issued a report and recommended a redistricting plan on October 18, 2001. The report noted the many constraints under which the committee operated, including population equality requirements, respecting racial minority representation, and the goal of keeping communities united. 2001 House Doc. No. 4700 at 4-6. Certain cities and towns could each naturally comprise a single district, and they were kept "united whenever possible." Id. at 5. "The process, however, [was] not an exact science. One community with a precinct or two too many [did] not necessarily abut another community in need of a precinct or two." Id. All of the recommended districts in the report were within the ± 5% standard, and the committee recognized that this standard allowed some latitude to keep certain communities united. Id. The committee recommended that Chelmsford be divided among four representative districts.
Prior to the present redistricting statute, the entirety of Chelmsford was combined with the town of Carlisle to comprise a single district. See St. 1993, c. 273, § 1. According to the 2000 census on which the challenged redistricting plan is based, the population of Chelmsford is 33,858. Because the ideal population size of each district is 39,682, the entire population of Chelmsford comprises more than 15% less than an ideal district.
On October 22, 2001, the committee's plan was debated by the full House, which adopted twenty-eight separate sets of amendments to the recommended districts, including eight sets of amendments proposed by Republicans. These amendments substantively changed a total of seventy-four districts — nearly one-half of the 160 districts in the plan. Chelmsford remained divided among four representative districts, although one of the districts was slightly modified by amendment. The House then adopted the amended plan by a roll call vote, with 128 representative voting in favor of the amendments and twenty-three voting against them. A substantial number of Republicans (eight of twenty-one) voted "Yea" and ten Democrats voted "Nay." The bill was passed by both the House and Senate and then signed by the Governor on November 8, 2001. St. 2001, c. 125.
The redistricting statute includes 116 districts that do not meet the art. 101 goal of avoiding combinations of political subdivisions. It divides Chelmsford among the following four districts: Second Middlesex district, with a population of 39,881; Fourteenth Middlesex district, with a population of 39,509; Sixteenth Middlesex district, with a population of 41,023; and Seventeenth Middlesex district, with a population of 41,363.
Chelmsford residents will comprise 27% of the Second Middlesex district, 20% of the Fourteenth Middlesex district, 28% of the Sixteenth Middlesex district, and only 9% of the Seventeenth Middlesex district.
4. The plaintiffs' alternative plan. The plaintiffs have proffered to the court an alternative plan that would divide Chelmsford between two of the four representative districts contemplated by the redistricting statute. The proposed redivision affects the population of each district, and creates a different level of deviation among all four. The maximum population deviation, computed by calculating the percentage deviation from the ideal district population figure of the largest and smallest of the four districts, is 5.04% under the alternative plan, as compared to 4.68% under the redistricting statute. The average population deviation of the four districts, computed by averaging the percentage that each district deviates from the ideal district population, is 2.81% in the alternative plan, as compared to 2.14% under the redistricting statute and 2.45% for the State as a whole.
5. Discussion.
a. Art. 101. The plaintiffs first argue that the redistricting plan embodied in St. 2001, c. 125, is unconstitutional because it violates art. 101's directive that the Legislature keep intact cities, towns, and counties "as nearly as may be." Although the plaintiffs have offered an alternative plan that more closely meets the territorial integrity requirement of art. 101, they "cannot prevail in this proceeding unless they establish beyond a reasonable doubt that it is impossible by any reasonable construction to interpret the [redistricting] statute in harmony with art. 101." Merriam v. Secretary of the Commonwealth, 375 Mass. 246, 263 (1978). In this regard, the plaintiff bears a heavy burden of proof and "every reasonable presumption must be made in favor of the constitutionality of the statute." Id.
As discussed in Mayor of Cambridge v. Secretary of the Commonwealth,ante 476, 484 (2002), we have adopted a two-step inquiry to determine whether a redistricting plan meets the territorial requirements of art. 101. The plaintiff must first demonstrate that "the constitutional requirement of equal representation `can be achieved . . . without dividing a municipality and [that] an entire municipality can be kept in one district without a "ripple effect" on other defined districts.'"Id., quoting Merriam v. Secretary of the Commonwealth, supra at 264 (Wilkins, J., dissenting). A "ripple effect" is created when fewer divisions of the town in question result in the increased division of neighboring towns and communities. Id., citing Brookline v. Secretary of the Commonwealth, 417 Mass. 406, 422 (1994) (alternative plan created "ripple effect" where it resulted in additional division of neighboring towns). If this burden is met, the plaintiff must next establish that the Legislature's plan is without justification. Id. The Legislature's proffered justification is entitled to a "high degree of deference." However, the Legislature's discretion is not without limit; the Massachusetts Constitution "prohibit[s] any departure from either [the territorial integrity or population equality requirement] that is unreasonable in the light of all the limitations imposed upon legislative action." Merriam v. Secretary of the Commonwealth, supra at 260, quotingAttorney Gen. v. Secretary of the Commonwealth, 306 Mass. 25, 31-34 (1940).
We assume, without deciding, as the Secretary has, that the plaintiffs have met the first prong by proffering a satisfactory alternative plan. Under that plan, Chelmsford would be divided between only two districts; the population of the reconfigured Second, Fourteenth, Sixteenth, and Seventeenth Middlesex districts proposed by the plaintiffs would each be well within the Federal population equality ± 5% standard; and there would be no identifiable adverse impact on other municipalities.
We therefore turn to the justification, that is, a closer degree of population equality. The plaintiffs' plan creates higher maximum and average deviations in population equality for the four districts in question. As described above, population equality, or the "one person one vote" construct, has been the driving force behind court and constitutional reform of the electoral districting process. Although the plaintiffs' alternative plan meets Federal equal protection requirements,Brown v. Thomson, 462 U.S. 835, 842 (1983) (less than 10% maximum deviation), art. 101's requirement that the Legislature achieve population equality "as nearly as may be" leaves the Legislature with considerable discretion to seek closer approximation to population equality than is required by the Federal equal protection clause. Certainly, our Constitution does not require that once the Federal equality of population requirement is met, the Legislature must maximize territorial integrity rather than seek to achieve closer population equality among the districts. As stated in Mayor of Cambridge v.Secretary of the Commonwealth, supra at 482, the two requirements of art. 101 are inherently contradictory.
While it is the responsibility of the Legislature to draw the redistricting plan, there is no definitive way to find the "best" such plan, because there is no single "best" way for the Legislature, let alone the courts, to determine which trade-offs between population equality, minority vote dilution, and territorial cohesion are most appropriate. The interaction of these factors is extraordinarily complex in the development of the Statewide plan, requiring an accumulation of difficult, interlocking judgments. In simplest terms, and as evident from the history of redistricting statutes enacted since the adoption of art. 101, if the Legislature wishes to more closely achieve the ideal of districts with equal populations, it must often create more districts that cross city, town, or county lines.
Minority vote dilution was not a factor directly implicated in creating the districts among which Chelmsford was divided, but was very much a factor in redistricting decisions Statewide.
There is no presumptive number of districts into which a city or town of a given size may be divided to which the Legislature must adhere. Nor has this court ever established a guideline above which efforts to achieve greater population equality may not intrude. It is the Legislature that has the authority and the discretion to determine "the extent to which either of these requirements shall yield to the other."Merriam v. Secretary of the Commonwealth, supra at 260, quoting Attorney Gen. v. Secretary of the Commonwealth, supra at 31-34.
To the extent Chelmsford's complaint is premised on four being too many divisions for a town of 33,858, its plight is hardly unique. A cursory review of the redistricting statute reveals many others similarly situated including:
City or Town Population Number of Districts
Bourne 18,721 3
Easton 22,299 3
Freetown 8,472 3
Middleborough 19,941 3
Oxford 13,352 3
Randolph 30,963 3
Walpole 22,824 4
Westborough 17,997 3
Article 101 offers no guidance on whether the appropriate price to be paid for one less district that unites a city and a town is an additional one-half percent, or 1%, or 2% average or maximum deviation from population equality. It is left to the Legislature to make judgments regarding these trade-offs.
It is in this context that the plaintiffs must carry their burden of demonstrating beyond a reasonable doubt that the departures they complain of are "unreasonable." We recognize that the plaintiffs' plan and the Legislature's plan differ only in relatively small degrees. However, the Legislature reasonably could have concluded that it was preferable to seek a closer approximation of population equality in the districts in the Chelmsford area, as part of a Statewide plan that sought to balance these conflicting objectives throughout, at the expense of dividing the town among four districts. This choice is related to the permissible, well-understood, legislative objective of achieving greater equality of population. That Chelmsford was consequently divided among four rather than two districts does not constitute a "substantial derogation" from the art. 101 requirement that territorial integrity be honored "as nearly as may be." If it did, any number of similar divisions made throughout the State (see note 12, supra) would be subject to the same criticism. Because the statute is entitled to "every reasonable presumption in favor of [its] constitutionality," id. at 263, we will not disturb the Legislature's determination.
It is not, as the dissent suggests, inconsistent with one of the Legislature's stated objectives to use the flexibility of the Federal ± 5% standard to try and keep some communities that were within that deviation range together. It is undisputed that Chelmsford's population did not fall within that range.
b. Political gerrymandering. The plaintiffs next argue that St. 2001, c. 125, constitutes a partisan political gerrymander in violation of the Federal equal protection clause. They assert that Chelmsford has been and is currently represented in the House of Representatives by a Republican, and that the Legislature's plan will destroy Chelmsford's Republican influence by dividing Chelmsford among districts that are currently represented by Democrats. We reject the plaintiffs' claim, as they have not proved that the Legislature's plan will lead to the disproportionately low representation of Republicans in the House of Representatives as a whole.
The United States Supreme Court, in Davis v. Bandemer, 478 U.S. 109, 127 (1986) (plurality opinion), held that to establish an equal protection violation based on partisan political gerrymandering, a plaintiff must demonstrate that the redistricting scheme has both a discriminatory intent and a discriminatory effect. See Terrazas v. Slagle, 821 F. Supp. 1162, 1172 (W.D.Tex. 1993). As to discriminatory effect, the Court noted that "the mere fact that a particular apportionment scheme makes it more difficult for a particular group in a particular district to elect the representatives of its choice does not render that scheme constitutionally infirm." Davis v. Bandemer,supra at 131. To find an equal protection violation, we must find "a history (actual or projected) of disproportionate results" and indications that a party has been denied fair representation. Id. at 139. "[A]n equal protection violation may be found only where the electoral system substantially disadvantages certain voters in their opportunity to influence the political process effectively." Id. at 133. The claim must be supported by evidence of "continued frustration of the will of a majority of the voters or effective denial to a minority of voters of a fair chance to influence the political process." Id.
Davis v. Bandemer, 478 U.S. 109, 127 (1986), was decided by a fragmented Court. A majority held that partisan political gerrymandering complaints were justiciable. However, the Court was divided as to the requirements for such a claim. Id. "When a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, `the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds. . . .'" Marks v. United States, 430 U.S. 188, 193 (1977), quoting Gregg v. Georgia, 428 U.S. 153, 169 n. 15 (1976) (opinion of Stewart, Powell, Stevens, JJ.). We apply the plurality opinion, which provides the narrowest grounds. Terrazas v. Slagle, 821 F. Supp. 1162, 1172 n. 12 (W.D.Tex. 1993).
The plaintiffs have not shown that the Legislature's redistricting plan has a discriminatory effect. Although they assert that Chelmsford voters have, in the past, played a part in electing a Republican to the House of Representatives, they have not demonstrated that the division of Chelmsford will result in a diminished Republican influence in either the relevant districts or the House of Representatives as a whole. Nor have the plaintiffs shown discriminatory intent on the part of the Legislature. To succeed on a political gerrymandering claim, a plaintiff must show that the Legislature has made a purposeful, concerted effort to disfavor a political party through the redistricting process. The plaintiffs have offered no evidence indicating that the Legislature's redistricting statute was the result of such efforts.
6. Conclusion. Judgment is to be entered for the Secretary of the Commonwealth declaring that the redistricting plan in St. 2001, c. 125, § 1, does not violate art. 101, and otherwise denying the relief requested by the plaintiffs.
So ordered.
According to the 2000 census on which the challenged redistricting plan is based, the population of the town of Chelmsford is 33,858. The ideal population size of each district (in order to achieve perfect population equality across all 160 districts) would be 39,682. Thus, the entire population of Chelmsford comprises less than a single district and would, ideally, be combined with a portion of one other municipality to create a district of the perfect size. I understand that such ideal redistricting is impossible, and that compromises on both population equality and territorial integrity of the districts must be made in order to obtain a redistricting plan that, as a whole, satisfies both objectives of art. 101, § 1, as amended by arts. 109, 117, and 119 of the Amendments to the Massachusetts Constitution.
The Legislature has split Chelmsford into four separate districts (where it was previously contained within a single district, St. 1993, c. 273, § 1), placing 10,943 Chelmsford residents in the Second Middlesex district (total district population of 39,881), 7,759 Chelmsford residents in the Fourteenth Middlesex district (total district population of 39,509), 11,407 Chelmsford residents in the Sixteenth Middlesex district (total district population of 41,023), and 3,749 Chelmsford residents in the Seventeenth Middlesex district (total district population of 41,363). Chelmsford residents thus will comprise less than one third of the population of each of those districts. Chelmsford's territorial integrity has not been honored, nor has the pragmatic purpose underlying territorial integrity been met by creating any district in which Chelmsford residents predominate or even approach a majority.
Chelmsford residents will comprise 27% of the Second Middlesex district, 20% of the Fourteenth Middlesex district, 28% of the Sixteenth Middlesex district, and only 9% of the Seventeenth Middlesex district.
The plaintiffs have submitted a proposed plan dividing Chelmsford into only two districts. The reconfiguration of the Second, Fourteenth, Sixteenth, and Seventeenth Middlesex districts proposed by the plaintiffs is virtually identical to the Legislature's plan in terms of population equality and has no adverse impact on other districts. The plaintiffs have, by their proposed plan, met their initial burden of showing that a plan that protects Chelmsford's territorial integrity can readily be formulated, and that such a plan can simultaneously satisfy the equally important art. 101 command of population equality, without causing undesirable "ripple effects" across other districts. See Brookline v. Secretary of the Commonwealth, 417 Mass. 406, 420-422 (1994). Where that burden has been met, the Legislature's plan violates art. 101 if its division of Chelmsford into more districts "is without justification."Ante at, citing Brookline v. Secretary of the Commonwealth, supra at 421.
Beyond significantly improving Chelmsford's territorial integrity, the plaintiffs' plan places only wards of the city of Lowell in the Seventeenth Middlesex. Thus, the plaintiffs' plan enhances the territorial integrity of Lowell as well.
The court today accepts as "justification" for this fragmenting of Chelmsford the claim that the Legislature's plan achieves "a closer degree of population equality," ante at, noting that "the Legislature . . . has the authority and the discretion to determine `the extent to which either of these requirements shall yield to the other.'" Ante at, quoting Merriam v. Secretary of the Commonwealth, 375 Mass. 246, 260 (1978). However, the claimed improvements in measurements of population equality are not only minuscule, but have been misapplied to the facts of this case in a fashion contrary to their use in other redistricting cases. Infinitesimal improvements in improperly calculated measures of population equality do not qualify as a legitimate excuse for doubling the number of divisions of a town that is itself smaller than a single district. Indeed, the population equality gains claimed are so trifling as to be mere coincidence, not the product of a legislative decision to give greater preference to such measurements of population equality at the expense of territorial integrity. Rather, this post hoc justification for fragmenting Chelmsford is contrary to the Legislature's stated preferences as to how the balance between population equality and territorial integrity was to be struck. While I recognize that considerable deference must be accorded to legislative judgment in the complex redistricting process, the justification now proffered is devoid of any statistical significance and contrary to the Legislature's expressed views on the subject.
The court's attempt to minimize the importance of territorial integrity by reference to the evolution of art. 101 and to the history of ever increasing divisions of municipalities over the course of sequential redistricting plans is contrary to the express wording of the constitutional mandate at issue. "[S]uch district[s] shall be formed, as nearly as may be, without uniting two counties or parts of two or more counties, two towns or parts of two or more towns, two cities or parts of two or more cities, or a city and a town, or parts of cities and towns, into one district" (emphasis added). Art. 101, § 1. The requirement of population equality is set forth in identical terms. The districts are to be formed "so that each representative will represent an equal number of inhabitants, as nearly as may be" (emphasis added). Id. Given that it is impossible to achieve both goals perfectly and simultaneously, the Legislature has the discretion to make compromises between the two as it configures the specific boundaries of the 160 districts. It does not have discretion to downgrade the importance of either of these constitutional mandates, which are of equal force and are required to meet the identical standard of "as nearly as may be." The suggestion that the Legislature may simply treat territorial integrity as less important than population equality is contrary to the Constitution.
The two sets of population figures compare as follows. For the Second Middlesex as configured in the redistricting statute, the district population is 39,881 (199 higher than the ideally sized district of 39,682), whereas the plaintiffs' alternative plan for that same district gives it a population of 38,978 (704 lower than the ideal). For the Fourteenth Middlesex as designed by the statute, the population is 39,509 (173 lower than the ideally sized district), whereas the plaintiffs' plan gives that district a population of 40,890 (1,208 higher than the ideal). For the Sixteenth Middlesex, the statute allocates a population of 41,023 (1,341 higher than the ideally sized district), whereas the plaintiffs' plan gives that district a population of 40,981 (1,299 higher than the ideal). Finally, for the Seventeenth Middlesex, the redistricting statute assigns a population of 41,363 (1,681 higher than the ideally sized district), whereas the plaintiffs' plan gives the district a population of 40,927 (1,245 higher than the ideal). Thus, the plaintiffs' plan is closer to ideal size for two districts (Sixteenth and Seventeenth Middlesex), while the Legislature's redistricting statute is closer to the ideal for the other two districts (Second and Fourteenth Middlesex). Moreover, the deviations in the plaintiffs' plan are all smaller than the largest deviations for these four districts in the redistricting statute.
Specifically, the largest deviation in the plaintiffs' plan is in the Sixteenth Middlesex, with a population of 40,981, or 1,299 higher than ideal. The redistricting statute has greater deviation from the ideal size in both the Sixteenth Middlesex (population 41,023, or 1,341 higher than ideal) and the Seventeenth Middlesex (population 41,363, or 1,681 higher than ideal). In percentage terms, the plaintiffs' plan for these four districts deviates from the ideal by a maximum of 3.27%, whereas the statute's provisions for these four districts has one district that is 4.24% higher than ideal and another district that is 3.38% higher than ideal.
The defendant argues, however, and the court today accepts as "justification," that two statistical computations, applied to just these four districts in isolation, make the Legislature's plan superior in terms of population equality. First, the defendant presents the calculation of the "average deviation" from the ideal across the four districts. Specifically, if we average the percentage by which the four districts as configured by the redistricting statute deviate from the ideally sized district, that "average deviation" is smaller than the "average deviation" calculated on the four districts in the plaintiffs' plan. Under the redistricting statute, the "average deviation" for the four districts is 2.14% (i.e., they are on average 2.14% off of the ideal district), whereas the "average deviation" for the same four districts under the plaintiffs' plan is 2.81%. The difference between these two averages is thus .67, or two-thirds of one percentage point.
The other statistical measurement put forth to demonstrate the allegedly superior population equality of the redistricting statute is referred to as "maximum" or "total deviation," which measures the difference between the largest and the smallest districts. Applying that measurement to just these four districts, the difference between the largest and smallest districts in the redistricting statute is 4.68%. Calculating that same measurement on the largest and smallest of these four districts in the plaintiffs' plan, the difference is 5.04%. The difference between the two plans on this measurement applied to these four districts is a difference of .36, approximately one-third of one percentage point.
That is, the largest district of these four, under the redistricting statute, is the Seventeenth Middlesex, at 4.24% higher than ideal, while the smallest district is the Fourteenth Middlesex, at .44% below ideal. The range between those two is thus 4.68%.
The largest district of these four, under the plaintiffs' plan, is the Sixteenth Middlesex, at 3.27% higher than ideal, while the smallest district is the Second Middlesex, at 1.77% below ideal. The range between those two is thus 5.04%.
Thus, the alleged superiority of the redistricting statute in terms of population equality, that ostensibly justifies splitting Chelmsford (and Lowell) further than otherwise necessary, is measured in fractions of a single percentage point. While the Legislature is to be accorded great deference in weighing these competing interests, such de minimis improvements in population equality should not qualify as adequate justification for substantial derogation from the equally important constitutional requirement that territorial integrity be honored "as nearly as may be." Art. 101, § 1.
Moreover, when we look at the proffered statistical measures applied to the two redistricting schemes as a whole, one of the claimed differences disappears entirely, and the other shrinks to near zero. The measurements of "average deviation" and "maximum deviation" are utilized in review of redistricting plans by applying them to the entirety of the plan being considered. All of the cases cited by the defendant in support of the use of these statistical measures apply them to the entirety of a districting scheme. See Abrams v. Johnson, 521 U.S. 74, 98-99 (1997) (applied to entirety of Georgia's congressional districts); Brown v. Thomson, 462 U.S. 835, 838 (1983) (applied to entirety of Wyoming's State representative districts); Karcher v. Daggett, 462 U.S. 725, 728 (1983) (entirety of New Jersey's congressional districts); White v. Weiser, 412 U.S. 783, 785 (1973) (entirety of Texas's congressional districts); White v. Regester, 412 U.S. 755, 761, 764 (1973) (Texas State representative districts);Gaffney v. Cummings, 412 U.S. 735, 737 (1973) (Connecticut State legislative districts). Not surprisingly, the defendant cites no authority for the proposition that these specific statistical measures are of any utility when applied to only a few districts within a much larger plan. A plan's adherence to population equality requirements cannot be assessed by reference to anything less than the plan as a whole. After all, what is being assessed is the plan itself, not just some subset of districts within that plan. Thus, we should consider what difference in these statistical measurements for the entire plan would result if the plaintiffs' alternative plan were to be adopted.
The measurement of "maximum deviation," i.e., the difference between the largest and smallest districts in the entire plan, is the same under the redistricting statute as it is under the plaintiffs' proposed alternative plan. Under either approach, the largest and smallest districts in the Commonwealth remain the same. Thus, viewing the plan as a whole, the maximum deviation measure of population equality is not any different under the redistricting statute than it is under the plaintiffs' plan, and that measurement, properly applied, provides no "justification" whatsoever for doubling the divisions of Chelmsford.
The largest district would still be the Twelfth Hampden district, at a population of 41,642, or 4.94% above the ideal district. The smallest district would still be the Twelfth Middlesex, at a population of 37,801, or 4.74% below the ideal district. The total deviation between those two extremes is thus 9.68% under either the redistricting statute or the plaintiffs' plan.
Applying the measurement of "average deviation" to each plan as a whole yields an infinitesimal difference between the two. Specifically, the average deviation under the redistricting statute is 2.45%. In other words, considering all 160 districts, the average percentage by which they deviate from the ideal district size is 2.45%. Making the changes proposed in the plaintiffs' alternative plan, the average deviation across the resulting 160 districts would be 2.47%, again meaning that, under that plan, the average percentage by which the districts over all would deviate from the ideal district size would be 2.47%. The difference between those two figures is a mere .02%, that is, two one-hundredths of one per cent. As illustration of the utterly de minimis quality of that figure, that magnitude of percentage difference, applied to the ideal district population of 39,682, amounts to an average difference of eight people per district. Given that the sizes of the various districts as configured by the statute differ by several thousand people, differences of only eight people per district are meaningless. The notion that Chelmsford's fragmentation into four districts is justified by a desire to improve the distance from ideal population size by a mere eight people per district is patently absurd.
The insignificance of this claimed superiority of the population equality provided by the Legislature's redistricting statute may also be illustrated by comparing the magnitude of deviations tolerated by the Legislature in its redistricting statute to the magnitude of the deviations proposed in the plaintiffs' plan. In the redistricting statute, the Legislature frequently accepted deviations from ideal population size far greater than the 1.77% to 3.37% deviations now proposed in the plaintiffs' plan. The greatest deviation from ideal size in the plaintiffs' plan is in their proposed Sixteenth Middlesex, with a population of 40,981 (or 1,299 higher than ideal). In the redistricting statute, there are twenty-four districts that are even larger than that (and thus even farther above the perfect district population of 39,682). Similarly, the statute has another twenty-six districts with a population that deviates by more than 1,299 below the ideal. Thus, there are fifty districts in the redistricting plan approved by the Legislature (or almost one third of the 160 districts) that are "worse" in terms of population equality than the "worst" of the deviations proposed in the plaintiffs' plan. The plaintiffs' plan is clearly well within what the Legislature viewed as an acceptable magnitude of deviation from population equality perfection, and there is no suggestion that the Legislature decided that territorial integrity had to yield in order to prevent deviations of the size presented by the plaintiffs' plan.
Two of the districts that are larger than the plaintiffs' largest proposed district are the statute's Sixteenth and Seventeenth Middlesex districts, which also include portions of Chelmsford. Indeed, there are five Middlesex districts in the statute that are larger than the plaintiffs' largest proposed district. The Legislature's willingness to tolerate population deviations larger than those proposed by the plaintiffs is not explained by reference to some particular need in some distant part of the Commonwealth. These are deviations the Legislature viewed as acceptable in various parts of Middlesex County, including within Chelmsford itself.
Four of those twenty-five districts are within Middlesex County. See note 8, supra.
The deviations in the plaintiffs' plan are also well within the deviations permitted under the Fourteenth Amendment. See Voinovich v.Quilter, 507 U.S. 146, 161 (1993), quoting Brown v. Thomson, 462 U.S. 835, 842-843 (1983). Both the statute and the plaintiffs' plan are below the 5% mark set by Federal precedent (no more than 5% above or 5% below the ideal population size, for a maximum total deviation of 10%) and, as discussed above, it is the plaintiffs' plan, not the redistricting statute, that keeps these four districts even further below that maximum permissible figure by almost a full percentage point (see note 4, supra). This is not a case where the population deviations from the ideal needed to maintain Chelmsford's territorial integrity would be at (or even anywhere near) the 5% maximum allowed. Rather, they are all comfortably within the acceptable range.
To the contrary, the Legislature expressly identified that its preference was to compromise population equality up to the maximum deviations permitted in order to preserve territorial integrity. As evidenced by the Joint Special Committee's Report, the Legislature adopted the 5% standard (tolerating deviations from ideal size of up to 5% in either direction) as the appropriate population equality goal, and recognized that the "latitude" of the 5% measure had the desirable effect of allowing it "to keep together communities, which might be a bit smaller, as well as communities, which are a bit larger." Joint Special Committee on the Redistricting and Reapportionment, Report Relative to a New Division of the Commonwealth into One Hundred and Sixty State Representative Districts, 2001 House Doc. No. 4700, at 5. There is no suggestion that the Legislature intended to split up communities in order to achieve imperceptible gains in "average population deviation." The Joint Committee's Report does not even mention the "average" deviation, let alone announce any goal of minimizing that "average." Rather, the Legislature announced its intention to exploit the "latitude" of the 5% standard to "keep together communities." That 5% standard accepts deviations from the ideal of up to 1,984 people per district, indicating that the Legislature had no desire to obsess over average differences on the order of eight people per district. In accordance with that standard, the Legislature accepted population deviations far larger than those in the plaintiffs' plan and affirmatively utilized them so as to preserve and enhance territorial integrity. That is precisely what the plaintiffs' plan does. As such, it is consistent with, not contrary to, the Legislature's view of how the competing goals of population equality and territorial integrity are to be balanced.
The court today attempts to avoid the clear import of the Joint Committee Report by suggesting that somehow this comment was limited to those municipalities that could, by themselves, comprise a single district. Ante at. Nothing suggests that the Joint Committee was espousing such a cramped view of the utility of the 5% standard. Indeed, in the entire plan, there are only seven districts that are entirely comprised of the entirety of a single community, and six of those have a population that is within a few hundred of the ideal population of 39,682. Only one such single community comprising a single district required the "latitude" of the 5% standard to remain as a single district, that being the Fourth Worcester district, which is comprised of the entirety of Leominster, with a population of 41,303. The Report's reference to the "latitude" of the 5% standard allowing it to "keep together communities," identifying it as "an important facet of the map-making process," does not refer to the configuration of only the Fourth Worcester district. Joint Special Committee on the Redistricting and Reapportionment, Report Relative to a New Division of the Commonwealth into One Hundred and Sixty Representative Districts, 2001 House Doc. No. 4700, at 5. Rather, the announced goal of "keep[ing] together communities" is a clear reference to the concept of territorial integrity, and the reference to the "latitude" of the population equality standard that allowed the Committee to "keep together communities" signals a clear intent to utilize that "latitude" in the interest of preserving territorial integrity. Id. The Joint Committee Report does not suggest that the 5% standard was only useful only to the extent that it allowed a perfect single municipality-single district match.
There is no justification for the Legislature's fragmenting of Chelmsford. We must show great deference to legislative judgment in the intensely political and highly complex matter of redistricting, but, in my view, the court today has improperly descended from deference to abdication. The eight person per district "gain" in average population deviation engendered by the Legislature's plan is so de minimis as to be nonexistent, and reflects only mathematical coincidence, not legislative design. Meanwhile, the Legislature's division of Chelmsford into four separate districts, with no district in which Chelmsford even comes close to a majority, deprives Chelmsford of any semblance of territorial integrity. The Legislature has broad discretion to balance the competing interests of population equality and territorial integrity. It does not have the discretion to abandon one of the mandates in art. 101 merely to achieve a couple hundredths of a per cent gain on one measure of the other competing mandate, nor did the Legislature indicate any intent to give any weight — let alone such inordinate weight — to that "average deviation" measure of population equality. Rather, the Legislature strived for, and the plaintiffs' plan maintains, a total maximum population deviation of 9.68%, within the Federal standard, and claimed that it was using the "latitude" of that standard to preserve territorial integrity. It may not now point to its coincidental two one-hundredths of a percentage point improvement on a measurement it never even mentioned as the justification for depriving the town of Chelmsford of anything resembling territorial integrity. The plaintiffs have shown, beyond a reasonable doubt, that there is no lawful justification for splitting Chelmsford into four districts and that the Legislature's plan therefore violates art. 101. I therefore respectfully dissent.
I agree with the court's opinion, ante at, that the plaintiffs have failed to establish their claim that the Legislature's plan constitutes a partisan political gerrymander in violation of equal protection guarantees, as they have failed to show that the plan "substantially disadvantages certain voters in their opportunity to influence the political process effectively" or that there has been "continued frustration of the will of a majority of the voters or effective denial to a minority of voters of a fair chance to influence the political process." Davis v. Bandemer, 478 U.S. 109, 133 (1986) (opinion of White, J.). It is apparent, however, that splitting up Chelmsford in the way that the Legislature has done places portions of Chelmsford — which has historically voted for a Republican representative — into each of four districts that have historically voted for Democrats as representatives, with Chelmsford now comprising only a minority percentage of each of those four districts. The present Republican incumbent from Chelmsford will be placed within a district that is dominated by an incumbent Democrat from Lowell, and that Lowell incumbent was successful in obtaining a further amendment to the Legislature's plan that served to increase his domination in the newly formed district. The apparent reason behind the Legislature's plan — the favoring of one incumbent over another based on party affiliation — is not a legally acceptable justification for splitting Chelmsford into four districts, and the defendant did not proffer such a justification.
The court today suggests that Chelmsford's loss of territorial integrity is somehow not a matter of concern because it is not the only small town that could have fit within a single district but was instead split among several different districts. Ante n. 12. Of course, none of the towns listed has made any challenge to the redistricting statute, and we therefore know nothing of the merits of any challenges that could theoretically have been brought. In particular, we have no way of knowing whether, as has been shown here, those towns' territorial integrity could have been better protected while simultaneously maintaining population equality and without causing any adverse ripple effects within neighboring districts and municipalities. In the absence of any challenges and any alternative plan, we must assume that the multiple divisions of those other towns was in fact necessary to satisfy some other legitimate goal. However, the plaintiffs here have met their burden of formulating an alternative plan that satisfies the tests laid out inBrookline v. Secretary of the Commonwealth, 417 Mass. 406, 420-421 (1994), quoting Merriam v. Secretary of the Commonwealth, 375 Mass. 246, 264 (1978) (Wilkins, J., dissenting). Having met their burden of proving that the division of Chelmsford into so many pieces is unjustified and therefore unconstitutional, it is no solace to Chelmsford that other towns have also been divided into many pieces. That fact should certainly not deprive Chelmsford of the relief to which it is entitled.