Six groups of petitioners from various towns in the state challenge the reapportionment of legislative districts by the General Assembly during the 1992 legislative session. Five of the groups contest various House districts, and the other group challenges a Senate district. We dismiss five of the petitions, and, with respect to the other petition, remand the House plan to the Legislature for revision and correction, if possible.

I. Legislative Redistricting

To comply wfith federal and state constitutional requirements that state legislatures afford substantially equal weight to citizens’ votes, the General Assembly is required to redraw legislative voting districts after each federal decennial census or after a state census taken for the purpose of redistricting. Vt. Const, ch. II, § 73; 17 V.S.A. § 1903(a). In forming representative and senatorial districts “which shall afford equality of representation, the General Assembly shall seek to maintain geographical compactness and contiguity and to adhere to boundaries of counties and other existing political subdivisions.” Vt. Const, ch. II, §§ 13, 18. Statutory criteria also require, “insofar as practicable,” that the districts be drawn consistent with the following policies: “(1) preservation of existing political subdivision lines; (2) recognition and maintenance of patterns of geography, social interaction, trade, political ties and common interests; [and] (3) use of compact and contiguous territory.” 17 V.S.A. § 1903(b).

Following the 1990 census, the Legislature reapportioned the House and Senate. See 1991, No. 116 (Adj. Sess.) (initial House districts); 1991, No. 147 (Adj. Sess.) (subdivided House districts and established Senate districts). The Legislature also amended the statutory procedures for redistricting the House in the same bill that redrew the initial House districts. 1991, No. 116 (Adj. Sess.), §§ 7-12. Under the new law, the bipartisan Legislative Apportionment Board, whose members do not serve in the General Assembly, must prepare a tentative redistricting proposal, consider the responsive recommendations of the municipal boards of civil authority, and then prepare a final *13proposal for dividing the state into initial districts for the election of the 150 representatives. 17 V.S.A. §§ 1905-1906. The final proposal is referred to the appropriate legislative committee, and eventually the General Assembly, which may accept or amend the proposal, or substitute another plan. Id. § 1906.

House districts may have no more than two members, Vt. Const, ch. II, § 13, but larger districts may be approved preliminarily subject to further subdivision. See 17 V.S.A. § 1906a(c). Similarly, two-member districts may be subdivided at that time. Id. § 1906a(b). The boards of civil authority within districts that are subdivided must prepare a proposal for drawing the internal lines within the districts, based on considerations of “incumbencies” in addition to the statutory criteria specified above. Id. §§ 1906b(b), (e) and 1906c(b), (c). The proposal is referred to the appropriate legislative committee, and eventually the General Assembly, which “shall” approve the House districts proposed by the town boards “if they are consistent with the.standards set forth” in the statutes. Id. §§ 1906b(e), (f) and 1906c(e), (f). If a majority of the town boards fail to agree to a subdivision proposal for a two-member House district, the Legislature “may divide the initial district into single-member representative districts.” Id. § 1906b(e). If a majority of the town boards fail to agree to a subdivision proposal for a House district with three or more representatives, the Legislature “shall divide the initial district into representative districts.” Id. § 1906c(e).1

Any five citizens may petition this Court, which has original and exclusive jurisdiction, for review of a final House or Senate plan. See id. § 1909(a). If this Court finds the plan in violation of constitutional or statutory requirements, it must forward its decision to the General Assembly and retain jurisdiction until *14the Legislature has approved a plan conforming to those requirements. Id. § 1909(e); see Vt. Const, ch. II, § 73 (authorizing Supreme Court to order reapportionment of legislative districts if Legislature fails to revise the districts as required).

In the present case, the Legislative Apportionment Board proposed a Senate and a House plan, neither of which was accepted by the Legislature. The legislative committees devised their own plans, which were adopted by the Legislature in most respects. The initial multimember districts created by Act 116 were subdivided by Act 147, resulting in a final House plan with an overall deviation of 17.6%.2 The Legislature approved a final Senate plan with an overall deviation of 16.4%. Five petitions challenged specific districts within the House plan, and one petition challenged the Senate plan. In all but one of the petitions, hearings were held before masters, who took testimony and made findings of fact. See 17 V.S.A. § 1909(d). We declined to order any interim relief, pending resolution of the petitions, and elections have proceeded under the new redistricting plans.

II. The Standard of Review

Redistricting is “primarily a matter for legislative consideration and determination.” In re Senate Bill 177, 130 Vt. 365, 371, 294 A.2d 657, 660 (1972). Accordingly, the redistricting plans approved by the General Assembly are presumed to be valid, In re Senate Bill 177, 130 Vt. 358, 361, 294 A.2d 653, 654 (1972), and there is “a heavy burden of proof on those who allege that a redistricting plan violates the Constitution.” Davis v. Bandemer, 478 U.S. 109, 185 (1986) (Powell, J., concurring in part and dissenting in part). Further, it is primarily the Legisla*15ture, not this Court, that must make the necessary compromises to effectuate state constitutional goals and statutory policies within the limitations imposed by federal law. See In re 1988 Legislative Apportionment of House, Senate, & Congressional Districts, 469 A.2d 819, 827 (Me. 1983) (hereinafter In re 1988 Legislative Apportionment). Accordingly, the Legislature must resolve the tension that exists between the one-person, one-vote requirement and state laws concerning the maintenance of compact and contiguous districts made up of communities with common interests. If a plan is consistent with the fundamental constitutional requirement that districts be drawn to afford equality of representation, we will return it to the Legislature only when there is no rational or legitimate basis for any deviations from other constitutional or statutory criteria. See Holmes v. Farmer, 475 A.2d 976, 986 (R.I. 1984).

We will not reject a redistricting plan simply because the petitioners have devised one that appears to satisfy constitutional and statutory requirements to a greater degree than the plan approved by the Legislature. See Gaffney v. Cummings, 412 U.S. 735, 750-51 (1973) (redistricting plan is not rendered unconstitutional simply because some “resourceful mind” has come up with a better one). Of course, the presentation of a plan that substantially improves on the one proposed by the Legislature may cast doubt on the legality of the Legislature’s plan. In re 1988 Legislative Apportionment, 469 A.2d at 828. The petitioners’ burden, however, is not to establish that some other preferable plan exists, but to demonstrate the absence of a rational or legitimate basis for the challenged plan’s failure to satisfy constitutional or statutory criteria. In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d 132, 136-37 (Pa. 1992); see In re Senate Bills 177 & 88, 132 Vt. 282, 290, 318 A.2d 157, 162 (1974) (“Whatever this Court may believe about the wisdom of an alternative solution, our testing of this legislative function must be confined to its constitutional and statutory propriety ....”); In re Reapportionment of Colorado General Assembly, 828 P.2d 185, 189 (Colo. 1992) (court’s role is to measure the redistricting plan against constitutional standards; the choice among alternative plans is for the redistricting commission, not the court).

In reviewing the petitions, we must consider not only the specific violations claimed, but also those claims within the con*16text of the entire plan, keeping in mind the difficulties in satisfying the various legal requirements statewide. In re Senate Bills 177 & 83, 132 Vt. at 289, 318 A.2d at 162; see In re Reapportionment Plan for Pennsylvania General Assembly, 442 A.2d 661, 668 (Pa. 1981). Nevertheless, once petitioners have shown that the State has failed to meet constitutional or statutory standards or policies with regard to a specific part of the plan, the State then has the burden to show that satisfying those requirements was impossible because of the impermissible effect it would have had on other districts.3 See, e.g., In re *17Reapportionment of Colorado General Assembly, 828 P.2d at 195-96 (redistricting commission’s statement that substantial equality of populations among districts required dividing a county did not “rise to the level of an adequate factual showing that less drastic alternatives could not have satisfied the equal population requirement of the Colorado Constitution”); In re Legislative Districting of General Assembly, 193 N.W.2d 784, 791 (Iowa 1972) (state failed to sustain its burden to show why legislature could not comply with constitution’s compactness requirement).

We recognize that at least one jurisdiction has required petitioners to submit an alternative plan that demonstrates that a challenged district could be reformed to meet constitutional or statutory criteria more effectively without contravening those requirements elsewhere in the state. In re 1983 Legislative Apportionment, 469 A.2d at 831. But considering that the Legislature, not the petitioners, is responsible for producing a valid plan, and that the Legislature and its staff are better equipped to devise alternative plans, we believe that basic notions of fairness require that the State show that a constitutionally or statutorily deficient plan cannot be improved without violating the same or other requirements elsewhere. Further, any alternative plan presented by petitioners would not be what it purported to be. It might serve to build petitioners’ case that the Legislature’s plan violated constitutional or statutory criteria, but, as a practical matter, it could not serve as a truly viable alternative plan. Petitioners’ alternative plan would invariably affect un*18challenged districts and could not be adopted without further review and consideration by the Legislature.

III. The Petitions

A. The Montgomery Petition

Petitioners from Montgomery challenge the Legislature’s establishment of a new House district composed of the town of Montgomery, which is in Franklin County, and four towns from Orleans County. Petitioners argue that the Legislature ignored Vermont constitutional and statutory law by joining the town of Montgomery in a district with four Orleans County towns with which it has little in common. According to petitioners, the Legislature gave no consideration to any constitutional or statutory criteria, aside from equality of population among districts. We agree that there is no indication that the Legislature considered all relevant constitutional and statutory criteria with respect to the establishment of this district. Accordingly, we send this matter back to the Legislature for further consideration.4

*19There is no doubt that the overriding duty of the legislature in establishing a redistricting plan is to assure substantial equality of population among districts. Reynolds v. Sims, 377 U.S. 533, 568 (1964) (“the Equal Protection Clause requires that the seats in both houses of a bicameral state legislature must be apportioned on a population basis”); see In re Senate Bills 177 & 83, 132 Vt. at 286, 318 A.2d at 160 (other considerations such as maintaining integrity of county lines must yield to that of equal representation). Our constitutional and statutory language makes this point clear. The constitution mandates that House and Senate districts “shall afford equality of representation,” whereas the constitution requires only that the Legislature “seek to maintain geographical compactness and contiguity and to adhere to boundaries of counties.” Vt. Const, ch. II, §§ 13,18. Statutory law requires the Legislature, insofar as practicable, to draw districts consistent with “policies” of preserving existing political lines, maintaining communities of interest, and using compact territory. 17 V.S.A. § 1903(b).

Nevertheless, state reapportionment laws “are not subject to the same strict [equal representation] standards applicable to reapportionment of congressional seats.” White v. Regester, 412 *20U.S. 755, 763 (1973). Early on, the Supreme Court acknowledged the practical impossibility of arranging state legislative districts “so that each one has an identical number of residents,” and required only “that a State make an honest and good faith effort to construct districts ... as nearly of equal population as is practicable.” Reynolds v. Sims, 377 U.S. at 577; see also Vt. Const, ch. II, § 73 (Legislature shall “maintain equality of representation among the respective districts as nearly as it is practicable”). Thus, “minor deviations from mathematical equality among state legislative districts are insufficient to make out a prima facie case of invidious discrimination under the Fourteenth Amendment so as to require justification by the State.” Gaffney v. Cummings, 412 U.S. at 745. As a general rule, an overall deviation under 10% within a redistricting plan is considered a minor deviation that does not require justification by some rational state policy. Brown v. Thomson, 462 U.S. 835, 842 (1983); see Gaffney v. Cummings, 412 U.S. at 745 (de minimis deviation does not violate equal protection clause whether it is considered alone or compared with another plan with lower deviations); In re Legislative Districting, 475 A.2d 428, 440 (Md. 1984) (plan with less than 10% deviation was “well within the permissible limits for state legislative districts” under federal constitution); Holmes v. Farmer, 475 A.2d at 988 (overall deviation of 5.4% is a minor deviation that is prima facie constitutional absent a showing of bad faith).

Accordingly, “nonnumerical”5 constitutional and statutory criteria are not rendered superfluous either by the primary concern for the one-person, one-vote requirement or by the flexibility the legislature has in meeting the nonnumerical criteria. See Schrage v. State Board of Elections, 430 N.E.2d 483, 486 (Ill. 1981) (dominant equal protection requirement should not be read as “death knell” to compactness requirement); Ater v. Keisling, 819 P.2d 296, 300, 303 (Or. 1991) (redistricting statute with criteria similar to those in Vermont statute viewed as giving legislature “latitude” in applying criteria). Indeed, this Court has a mandate to retain jurisdiction over redistricting *21until the Legislature produces a plan that conforms to “all constitutional and statutory requirements.” 17 V.S.A. § 1909(e) (emphasis added).

All of the nonnumerical criteria at issue here — the maintenance of political lines, the use of compact and contiguous territories, and the preservation of communities with common interests — are not only important but are related to one another in that they share the common purpose of assuring more effective representation. See Gaffney v. Cummings, 412 U.S. at 748-49 (fair and effective representation “does not depend solely on mathematical equality among district populations”); Reynolds v. Sims, 377 U.S. at 578-79 (“Indiscriminate districting, without any regard for political subdivision or natural or historical boundary lines, may be little more than an open invitation to partisan gerrymandering.”). Thus, the preservation of town and county boundaries is important “because the sense of community derived from established governmental units tends to foster effective representation.” Carstens v. Lamm, 543 F. Supp. 68, 88 (D. Colo. 1982). Local governmental units have various responsibilities incident to the operation of state government in a wide range of areas, including the court system, law enforcement, education, mental health, taxation, and transportation. Consequently, unnecessary fragmentation of these units limits the ability of local constituencies to organize effectively and increases voter confusion and isolation. Id.; see Davenport v. Apportionment Commission, 304 A.2d 736, 745 (N.J. Super. Ct. 1973) (“‘citizens of each county have a community of interest by virtue of their common responsibility to provide for public needs’”) (quoting Jackman v. Bodine, 205 A.2d 713, 718 (N.J. 1964)). Voters in a community are less effectively represented when their elected representative’s principal constituency lies outside their community and has interests different from their own. See In re Reapportionment Plan for Pennsylvania General Assembly, 442 A.2d at 669 (Nix, J., dissenting). These considerations are particularly relevant in this state, which has a long history of preserving the independence and integrity of local government. See In re Senate Bills 177 & 83, 132 Vt. at 287, 318 A.2d at 160.

*22Similarly, compactness and contiguity requirements ultimately concern “the ability of citizens to relate to each other and their representatives and ... the ability of representatives to relate effectively to their constituency.” Wilson v. Eu, 823 P.2d 545, 553, 4 Cal. Rptr. 2d 379, 387 (1992). These relationships are fostered through shared interests and membership in a political community. Id.; see In re Reapportionment Plan for Pennsylvania General Assembly, 442 A.2d at 669 (Nix, J., dissenting) (requirements of contiguity and compactness go beyond geographical concern and embrace concept of homogeneity of district). They are undermined, however, when geographic barriers that severely limit communication and transportation within proposed districts are ignored. See Wilson v. Eu, 823 P.2d at 574, 4 Cal. Rptr. at 400 (Appendix I). Thus, the statutory criteria contained in § 1903(b), including “recognition and maintenance of patterns of geography, social interaction, trade, political ties and common interests,” are an implementation and extension of our constitutional requirements that the Legislature “seek to maintain geographical compactness and contiguity and to adhere to boundaries of counties and other existing political subdivisions.” Vt. Const, ch. II, §§ 13, 18.

We now examine the findings of the master, which are not in dispute. The towns of Jay, Troy, Westfield, Lowell, and Montgomery comprise the challenged district, Orleans-Franklin 1, which has a population of 3,829 and an overall deviation of 2.05%. The northern spine of the Green Mountains divides Montgomery from the Orleans County towns, affecting transportation and communication between them, particularly in the winter. Although there are two roads that run east from Montgomery over the spine, one is a seasonal road and the other is difficult to travel in the winter. Because of the natural barriers to the east, Montgomery’s economic, social, educational and political ties have always faced away from the mountains toward Richford, Enosburg and other Franklin County towns. For example, trade and commerce generally moves westward from Montgomery, children go to high schools in towns in the school district to the west, social and sporting events generally take place with towns to the west, serious health problems are dealt with in hospitals to the west, and except for a few Montgomery *23residents who work at the Jay Peak ski area, the work force moves in a westerly direction. Because of the lack of shared interests between towns in the two counties, the weekly and daily newspapers of Franklin and Orleans counties generally do not cover local events outside their own counties.

In sum, these findings, and other evidence presented at the hearing before the master, indicate that Montgomery was placed in a district with towns from a different county separated by a mountain range that greatly limits transportation and communication — and therefore, social, political and economic interaction — between the two areas.6 The only explanation for the placement provided by the State is that the Orleans-Franklin 1 district did not present an equal protection problem. When asked why Montgomery was joined in a district with Orleans County towns, the minority leader of the House Government Operations Committee stated that the Committee “felt that there was a connection with Jay Peak and that historically and in many ways there is a strong connection in Montgomery and Franklin County. But we were bound by numbers(Emphasis added.) He was unable to provide details or support for any connection between the two communities.7 Although the *24minority leader stated how difficult it was to “put together districts around the State” without exceeding a 16.4% maximum deviation,8 he never indicated that the Committee considered nonnumerical criteria but was unable to come up with a plan that satisfied both equal population concerns and the other criteria. Rather, he testified that there were numerous ways the Committee could have gone, and that there was no specific vote on Montgomery. Further, he conceded that the redistricting was “based primarily upon numbers and that it was driven by arithmetic.”

In short, there is nothing in the record before us, including the State’s brief, indicating either that the Legislative Apportionment Board or the Committee considered nonnumerical criteria with respect to the Orleans-Franklin 1 district, or that the Board or Committee could not produce a plan that adhered to all criteria with regard to that district.9 Cf. State ex rel. Lockert v. Crowell, 656 S.W.2d 836, 838 (Tenn. 1983) (court affirmed lower court decision rejecting redistricting plans that were drafted with sole objective of obtaining low percentages of total *25deviation, with no effort made to consider other nonnumerical criteria); compare In re Reapportionment of Colorado General Assembly, 828 P.2d at 195-96 (plan returned to redistricting commission for reconsideration and resubmission because plan joined in one district towns of different counties separated by mountains), with In re Reapportionment of Colorado General Assembly, 828 P.2d 213, 216 (Colo. 1992) (resubmitted plan approved, though similar to the one rejected earlier, because commission provided court with sufficient basis for judicial review of its actions). Instead, the State argues that the equal population requirement is paramount and contends that putting Montgomery in a Franklin district, without making any other changes, would violate that requirement. Although we give great deference to the Legislature’s redistricting decisions, such decisions are not foreclosed from further review when the Legislature makes only a general reference to equal population and does not justify or explain why a district fails to satisfy any of the nonnumerical criteria. To hold otherwise would be to render meaningless those nonnumerical criteria.

Petitioners argue that placing Montgomery with other Franklin County towns in the Franklin 2 district would satisfy all constitutional and statutory criteria. The master found that placing Montgomery in the Franklin 2 district, and not otherwise changing the House plan, would improve the deviation in that district from -8.3% to 2.3%. The deviation in the Orleans-Franklin 1 district, however, would widen from 2.1% to -19.9%, thereby creating an overall deviation within the House plan of 28.8%, approximately 11.2% greater than that of the current plan. Petitioners contend that such a deviation, if justified by legitimate state policies, such as maintaining county lines or communities of common interest, would pass constitutional muster under the equal protection clause. We disagree; the case law provides no such assurance. The United States Supreme Court, in permitting an overall deviation of 16.4% in one state redistricting plan based on rational state objectives, warned that “this percentage may well approach tolerable limits.” Mahan v. Howell, 410 U.S. 315, 329-30 (1973); see State v. Crowell, 656 S.W.2d at 839 (overall deviation of 21% would be unacceptable under Mahan). Indeed, this Court has already rejected on equal protection grounds a Senate plan that had a maximum *26deviation of 25.3%. In re Senate Bill 177, 130 Vt. at 370, 294 A.2d at 659-60.

Petitioners contend, however, that recent federal case law allows much greater deviation when justified by rational state objectives, but the cases they cite do not support their argument. In Swann v. Adams, 385 U.S. 440, 444-46 (1967), the Court actually rejected a plan with a maximum deviation of 26%, see Mahan v. Howell, 410 U.S. at 328, and only suggested by negative inference that the State’s presentation of rational state policies could have changed the result. In Brown v. Thomson, 462 U.S. at 839, 846, the Court avoided ruling on the constitutionality of a Wyoming redistricting plan with a -89% maximum deviation that resulted from giving a representative to a sparsely populated county. Because of the limited review in that case and Wyoming’s unique situation, the Brown opinion has been considered an aberration with little precedential value. See id. at 850 (Brennan, J., dissenting) (“it is worth stressing how extraordinarily narrow [the holding] is, and how empty of likely precedential value”); State v. Crowell, 656 S.W.2d at 840 (“the unusual single deviation in Brown may well be inapplicable elsewhere”); see also New York City Board of Estimate v. Morris, 489 U.S. 688, 702 (1989) (citing, among other cases, Brown v. Thomson, Court noted that “no case of ours has indicated that a deviation of some 78% could ever be justified”). Assuming that the Legislature’s only choice was to place Montgomery in the Orleans-Franklin 1 district or the Franklin 2 district, we could not conclude that its decision to place Montgomery with towns in Orleans County was irrational or illegitimate.

Citing a plan appended to another petition, petitioners also argue that the Legislature could have kept Montgomery with Franklin County towns and achieved minimal population deviations in the overall plan. Because the plan has never been presented as part of this case,10 we do not consider it here. Regardless of the merits of the cited alternative plan, peti*27tioners contend that because they have shown that the statutory criteria have not been adhered to, it is up to the Legislature, not them or this Court, to come up with an alternative plan that satisfies both equal population and the other criteria. The apparent flaw in this argument is that, unless there is some showing that such an alternative plan can be produced, it is unclear whether the Legislature has ignored its duty to consider the nonnumerical criteria. Nevertheless, we believe that further legislative consideration is appropriate in this instance, where (1) petitioners have shown that none11 of the nonnumerical statutory or constitutional criteria were adhered to with regard to the Town of Montgomery; (2) the State has failed to provide any rational reason for Montgomery’s placement, other than a general reference to equal population; and (3) the State has not shown, by evidence or affidavit, that an alternative plan satisfying the various constitutional and statutory criteria could not be produced.

As we have noted, redistricting is a legislative function, and equal population is the overriding objective. We have a duty, however, to review redistricting plans to assure that the Legislature has given full consideration to all constitutional and statutory criteria. We cannot allow equal population to be elevated from an overriding objective to the sole consideration. The nonnumerical criteria are neither superfluous nor unimportant. *28They assure that our citizens obtain effective representation by being placed in districts comprised of communities with shared interests.

Petitioners’ burden is not to establish that there exists a preferable alternative, but to demonstrate that a plan does not meet constitutional or statutory standards. In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d at 136. Once petitioners have made such a showing, the State must demonstrate why a better alternative was not possible, or that an alternative plan submitted by the petitioners is flawed. See Davenport v. Apportionment Commission, 304 A.2d at 746-47 (where reapportionment commission failed to show whether, or to what extent, compactness was considered, court remanded plan for commission to determine whether it would be possible to devise an alternative plan that considered state constitutional criteria); State v. Crowell, 656 S.W.2d at 843 (house reapportionment plan held to be unconstitutional because legislature failed to justify substantial crossing of county lines); cf. In re Legislative Districting of General Assembly, 193 N.W.2d at 791 (legislature failed to sustain its burden to show why it could not comply with state constitutional compactness requirement). But see In re 1983 Legislative Apportionment, 469 A.2d at 831 (petitioners must show that district could be improved without creating constitutional violations elsewhere in state). Accordingly, we return the House plan to the Legislature for consideration of the nonnumerical criteria with respect to the Orleans-Franklin 1 district and determination of whether Montgomery can be placed in a district that satisfies those criteria without creating constitutional or statutory violations elsewhere in the state.

B. The Springfield Petition

Petitioners from Springfield challenge various aspects of Act 116 and Act 147. They argue that (1) the House plan violates the equal protection clause because it has an overall deviation greater than 10%, and the State has failed to show that the deviation was the necessary result of the implementation of legitimate state policies; (2) their initial district violates constitutional and statutory criteria because it crosses a county line and lacks contiguity, compactness, and common interests; (3) the subdivided districts created by the Legislature resulted *29from political gerrymandering rather than the application of the relevant constitutional and statutory criteria; (4) the incumbency criterion for subdividing multimember districts violates the Vermont Constitution; (5) Act 116 violates equal protection by giving each town board of civil authority, regardless of population, an equal vote in subdividing multimember districts; and (6) the Legislature violated statutory procedure when it subdivided their initial district. We reject each of these arguments, and dismiss the petition.

Petitioners first argue that the State failed to adequately justify the House plan’s overall deviation of 17.6%. We agree with the State that petitioners have waived their right to raise this issue at this juncture because it was not raised in their petition or at the hearing before the master. See In te Mullestein, 148 Vt. 170, 175, 531 A.2d 890, 893 (1987) (absent extraordinary circumstances, constitutional issues not raised before trial court are waived on appeal). Petitioners contend that the issue was adequately raised by a request for relief in their petition — “Declare Act 116 unconstitutional under the Equal Protection Clause of the 14th Amendment of the United States Constitution” — and by unspecified evidence presented at the master’s hearing.12 But the prayer for relief was one of fifteen various requests for relief in a thirty-page petition that never specifically challenged the overall deviation of the House plan. Further, a challenge based on the plan’s overall deviation was not specifically addressed at the master’s hearing. It is not enough that there was evidence before the master that could have supported this claim. The point is that the State was not given adequate notice of the claim. Further, the fact that this Court has original jurisdiction is not significant. The purpose of the master is to take testimony and make findings for this Court. See 17 V.S.A. § 1909(d). Because the State was not on notice that petitioners were challenging the overall deviation of the plan, it had no opportunity to create a record before the master that included justification for the deviation.

*30Next, petitioners argue that inclusion of Springfield in a district with three Windham County towns violated constitutional and statutory criteria requiring the Legislature to create compact and contiguous districts that, if possible, maintain county lines and common interests. We reject this argument on procedural grounds. First, petitioners are challenging the initial Windsor-Windham 1 district, which no longer exists as a “representative” district. See id. § 1903(b) (stating standards for the creation of representative and senatorial districts). We are aware that the town boards of civil authority are directed to consider the § 1903 standards when recommending proposals of initial districts to the Legislative Apportionment Board, see id. § 1905, but the fact of the matter is that the initial WindsorWindham 1 district has been subdivided and does not exist as a representative district. Therefore, petitioners cannot complain that it is not compact or that towns within it have no common interests.

Second, even if we accepted petitioners’ argument as a challenge to the subdivided Windsor-Windham 1 district, we agree with the State that petitioners have no standing to raise this issue because they do not reside in the challenged district. Petitioners correctly point out that many of their arguments entail a broader challenge, foreclosing dismissal of their petition on grounds of standing. But that is not the case with this argument. Petitioners all reside in the Windsor 6 district, yet they argue that Windsor-Windham 1 is invalid because it crosses county lines, is not compact or contiguous, and is not composed of towns with common interests. We address their broader political gerrymandering claims below, but we do not accept this narrow challenge to a district in which none of the petitioners reside. ■

Finally, even if we considered petitioners’ challenge of Windsor-Windham. 1 on the nonnumerical criteria, we would reject it because petitioners have failed to meet their burden of showing that the Legislature had no rational or legitimate basis for creating the district. Petitioners’ principal argument is that the nonnumerical. constitutional and statutory criteria would be better met by placing Springfield with other towns. The question is not whether there is a better alternative plan, however, but whether the Legislature’s plan violates the legal standards.

*31We find no such violation. The boot-shaped district clearly does not violate principles of compactness and contiguity. See, e.g., Schrage v. State Board of Elections, 430 N.E.2d at 486-87 (citing cases involving noncompact districts and describing one challenged district as “tortured” and “extremely elongated”); In re Legislative Districting, 475 A.2d at 443-44 (describing odd-shaped districts that might pose compactness problem, and noting that presentation of alternative only begs the question of whether a district meets constitutional or statutory requirements). Although the master found that the town of Springfield was not contiguous with the towns of Windham and Grafton, he did not find that the district lacked contiguity. On the contrary, he found that all four towns in the district shared at least one common border. Further, petitioners presented virtually no evidence to show a lack of common interests among the towns in Windsor-Windham 1. Indeed, the master found that there is a network of state highways connecting the towns. The only evidence presented by petitioners showing that Windsor-Windham 1 does not meet a nonnumerical criterion is that the district breaches a county line. That alone, however, does not prove that there was no rational or legitimate basis for creation of the district. See In re Senate Bills 177 & 83, 132 Vt. at 289, 318 A.2d at 162 (crossing county lines does not disqualify plan, given difficulties of reaching an acceptable result statewide); In re Reapportionment Plan for Pennsylvania General Assembly, 442 A.2d at 668 (crossing county lines is unavoidable).

Petitioners also argue that the subdivision of Windsor 6 violated the requirement that the Legislature seek to adhere to boundaries of existing political subdivisions because the district line within Springfield splits too many streets, relies too much on boundaries of census blocks, and, at times, ignores all natural boundaries. Petitioners, however, have failed to indicate which political subdivisions were ignored, or to refute adequately the State’s contention that the line was drawn to separate Springfield’s most urban areas from neighborhoods that would have more in common with the relatively rural town of Rockingham. We cannot conclude that there was no rational and legitimate basis for the internal line drawn within Springfield.

*32Petitioners’ next contention is that the subdivision of Windsor-Windham 1 amounted to political gerrymandering in favor of the Democratic Party, in violation of the compactness and contiguity requirements of Chapter II, § 13 and Chapter I, Article 7, the Common Benefits Clause of the Vermont Constitution. They rely chiefly on hearsay testimony to the effect that the Speaker of the House indicated that he wanted internal lines drawn in Springfield to protect certain Democratic incumbents. According to petitioners, this goal was accomplished by forming districts that would pit a Republican incumbent against two unbeatable Democratic incumbents rather than against a more vulnerable Democratic incumbent. This argument is without merit. Political considerations are an inevitable component of redistricting and are not per se improper. In re Reapportionment of Colorado General Assembly, 828 P.2d at 199; see Davis v. Bandemer, 478 U.S. at 129 (“As long as redistricting is done by a legislature, it should not be very difficult to prove that the likely political consequences of the reapportionment were intended.”); Gaffney v. Cummings, 412 U.S. at 753 (“districting inevitably has and is intended to have substantial political consequences”); see also In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d at 142 (no individual has a right to run for a particular office, nor do citizens have a right to vote for a specific individual).

In challenging the formation of a district, the fundamental question is whether constitutional or statutory criteria have been violated, not whether the legislators intended to obtain some political advantage. See Republican Party of Virginia v. Wilder, 774 F. Supp. 400, 404 (W.D. Va. 1991) (even if petitioners showed that “district boundaries were drawn solely for partisan ends,” they must prove an actual discriminatory effect) (emphasis in original). Because we have found that neither of the challenged districts violates constitutional or statutory criteria, we need not consider the motives of the legislators. In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d at 147 (because redistricting plan does not violate state constitution on its face, court need not resolve whether members of the commission can be compelled to disclose underlying motives); see People ex rel. Burris v. Ryan, 588 N.E.2d 1033, 1038 (Ill. 1992) (“A map that is politically unacceptable to one political party is not, for that reason, legally *33unacceptable.”) (emphasis in original). If we were to accept gerrymandering claims based solely on a showing of an intent to favor a political party, we would commit the responsibility for redistricting to the courts rather than to the Legislature.

Next, petitioners challenge two of the procedures added by Act 116 for subdividing multimember districts. First, petitioners argue that Act 116 violates Chapter II, § 13 and Chapter I, Article 7 of the Vermont Constitution by requiring boards of civil authority to consider “incumbencies” when creating subdivision proposals for multimember districts. See 17 V.S.A. §§ 1906b(c)(4) and 1906c(c)(4). The thrust of petitioners’ argument is that protecting incumbents is just another form of political gerrymandering, which the criteria of Chapter II, § 13 are meant to prevent. We disagree that consideration of incumbents in any manner constitutes a per se violation of the Vermont Constitution. As long as constitutional and statutory criteria regarding redistricting are adhered to, including those criteria contained in Chapter II, § 13, creating districts to avoid contests between incumbents is a legitimate consideration that may justify minor deviations from equal representation. Karcher v. Daggett, 462 U.S. 725, 740 (1983); see In re Legislative Districting of General Assembly, 193 N.W.2d at 790 (minimizing contests between incumbents is not impermissible consideration unless it results in districts lacking population equality or compactness). The record indicates that the incumbency criterion was narrowly interpreted to mean avoiding alternatives that create contests between incumbents,13 which is the limited construction we give it. As noted above, there is no showing that consideration of this criterion created a district that violated any constitutional or statutory criteria.

Regarding the Common Benefits Clause,14 petitioners have not indicated what part of, or how, the community has been dis*34advantaged. Cf. In re One Church Street, 152 Vt. 260, 264, 565 A.2d 1349, 1351 (1989) (construing State v. Ludlow Supermarkets, Inc., 141 Vt. 261, 448 A.2d 791 (1982)) (supporters of Sunday closing law could not claim that large businesses compelled to close on Sundays were included in the general class of beneficiaries, or that the detriment they incurred was only incidental). Petitioners complain that the incumbency criterion is a form of gerrymandering, yet, under the narrow construction we have given it, it can reduce the potential for gerrymandering by making it more difficult to pit a targeted incumbent against another incumbent who is more likely to prevail. .See State v. Ludlow Supermarkets, Inc., 141 Vt. at 269, 448 A.2d at 795 (preferential legislation must further a goal independent of the preference awarded). Although two incumbents had to be pitted against one another in this case, we believe that in cases where it is possible to avoid this result, the criterion will prevent gerrymandering. We conclude that the purpose of the criterion is reasonably related to a legitimate state interest. See Town of Sandgate v. Colehamer, 156 Vt. 77, 88, 589 A.2d 1205, 1211 (1990) (test of constitutionality is whether law’s purpose is reasonably related to promoting valid state interest).

The second provision of Act 116 to which petitioners object concerns the procedure by which town boards of civil authority share responsibility for preparing a proposal for subdivision of multimember districts. After enactment of a final plan for initial districts, boards of towns within multimember districts are directed to meet and prepare a proposal for division of the district. 17 V.S.A. §§ 1906b(b) and 1906c(b). Each town, regardless of population, has one vote on accepting or rejecting any proposal, and any town may veto a proposal that draws internal lines within that town. Id. §§ 1906b(b) and 1906c(b). If a majority of the towns agree on a proposal, and it is consistent with the statutory standards, the Legislature must accept it. Id. §§ 1906b(f) and 1906c(f). On the other hand, if the boards are unable to obtain a majority vote on a proposed subdivision, the Legislature must subdivide the district. Id. §§ 1906b(e) and 1906c(e). Petitioners contend that giving each town one vote, regardless of population, offends the Equal Protection Clause of the Fourteenth Amendment to the United States Constitution. We disagree.

*35Petitioners rely heavily on two cases, which stand for the proposition that under certain circumstances the makeup of local governing boards is subject to the one-person, one-vote principle of Reynolds v. Sims, 377 U.S. 533 (1964). See New York City Board of Estimate v. Morris, 489 U.S. 688 (1989); Kelleher v. Southeastern Regional Vocational Technical High School District, 806 F.2d 9 (1st Cir. 1986). Petitioners correctly point out that equal representation applies to local governing bodies in situations where the members of the local body are elected officials, and the body exercises a wide range of local governmental functions. Regarding the second criterion, the Court in Board of Estimate listed a litany of significant functions common to municipal governments that are performed by ■New York’s Board of Estimate — including calculating tax rates, zoning, managing city property, fixing salaries of city employees, granting city contracts, and sharing responsibility for formulation of the city budget — and concluded that these ‘“powers are general enough and have sufficient impact throughout the district’ to require that elections to the body comply with equal protection strictures.” 489 U.S. at 696 (quoting Hadley v. Junior College District of Metropolitan Kansas City, 397 U.S. 50, 54 (1970)); see Kelleher, 806 F.2d at 10-11 & n.7 (court lists wide range of powers given to regional school district committee that brings the committee within the one-person, one-vote principle).

According to petitioners, the second criterion is satisfied here because the boards of civil authority of each town perform many functions common to municipal governments. Petitioners miss the point, however. They are challenging the voting procedure of a “superboard” created by combining the boards of each town in a multimember district, not of an individual town board. This “superboard” does not perform a wide range of local governmental functions. Rather, the Legislature has delegated an aspect of redistricting, a legislative function, to this “superboard,” perhaps in deference to Vermont’s unique history as a state formed by an association of preexisting independent towns. Thus, rather than performing a wide range of functions common to local government, the “superboard” performs an extremely narrow role within a legislative function. There is no equal protection violation.

*36Assuming the constitutionality of § 1906c, petitioners argue that the Legislature violated that section by adopting its own subdivision of Windsor-Windham 1 despite the fact that a majority of the district’s boards of civil authority agreed upon their own subdivision. We cannot agree that this fact alone is sufficient to establish a violation of § 1906c. On or before April 1 of an election year, a majority of a district’s boards of civil authority must present a proposal for subdividing the district. 17 V.S.A. § 1906c(e). On March 30,1992, Springfield’s proposal for subdividing Windsor-Windham 1 was transmitted to the clerk of the House. The proposal required subdivision lines within both Springfield and Rockingham, but no internal lines were drawn through Rockingham. The House Government Appropriations Committee held a public hearing on April 2, 1992, in which it learned or was aware that Springfield, Grafton, and Windham supported the Springfield plan, but Rockingham objected to being subdivided. The Legislature was entitled to subdivide Windsor-Windham 1 because the Springfield plan required that Rockingham be divided, and Rockingham had a right to veto a plan that drew internal lines within its boundaries. Id. § 1906c(b).

We also reject petitioners’ argument that the Legislature’s subdivision favored only Democrats. Because of the decline in the population of the area, any subdivision plan would have had to pit the Republican incumbent against Democratic incumbents. We are not concerned with speculation as to which incumbent had the best chance to prevail over which incumbent in the other party. In this situation, one of the political parties had to be in a presumably favored position. We find no violation of any statutory or constitutional criteria.

C. The Berlin Petition

Petitioners from Berlin challenge the subdivision of an initial three-member district, comprised of the town of Berlin and the city of Barre, into a one-member district completely within Barre City, and a two-member district that included the town of Berlin and a portion of Barre City. The gist of petitioners’ argument is that Berlin should have been placed in a .one-member district with a small part of Barre so as to assure that at least one of the three representatives of the initial district would represent the interest of Berlin’s rural, small town voters.

*37Petitioners acknowledge that, in any scenario, residents of Barre and Berlin would have to be joined in a representative district. They believe they are entitled to be part of a one-member district composed of approximately 1500 Barre residents and 2500 Berlin residents. They point out that during the last decade the population of Berlin increased from 2454 residents to 2561 residents, while Barre’s population fell from 9824 residents to 9482 residents. Thus, Barre has a population that would support approximately two and one-half representatives, while Berlin’s population would support approximately two-thirds of a representative. Petitioners argue that, given these numbers and the fact that the residents of their rural community have virtually nothing in common with Barre residents, it is fundamentally unfair to disenfranchise them by placing them in a two-member district with a majority of Barre residents. We disagree.

The residents of Berlin are not disenfranchised simply because they make up a minority of their two-member representative district. Even in situations involving racial or political groups, proportional representation is not constitutionally required. See Davis v. Bandemer, 478 U.S. at 132 (one cannot presume that the winning candidate will entirely ignore the voters who supported the losing candidate; “a group’s electoral power is not unconstitutionally diminished by the simple fact of an apportionment scheme that makes winning elections more difficult”). Members of a group are disenfranchised only when they are denied an opportunity to effectively influence the election results by securing the attention of the winning candidate. Id. at 132-33. There has been no such showing here.

Ironically, petitioners acknowledge that approximately 1500 Barre residents would have to be joined with Berlin residents in any single-member district. They argue that those 1500 will not be disenfranchised because their interests will be represented by the other two Barre representatives. This argument is untenable. If the lack of common interests would preclude placing 5000 Barre residents with Berlin, it would also preclude placing 1500 Barre residents with Berlin.

Stripped of the disenfranchisement aspect of their claim, petitioners are left with the argument that placing Berlin and Barre residents in the same representative district violates *38Vermont’s statutory criteria because the two communities have no common interests.15 The master found that Barre and Berlin are in separate school districts, that the two towns do not share any essential governmental services, that they have distinct types of governmental structures, resulting in a much higher tax rate in Barre, and that, unlike Berlin, Barre has an urban residential pattern with a relatively high population density.

On the other hand, the master found that there is a network of roads connecting the two communities, that both communities have large commercial areas which serve residents from both towns, that many residents from each community work in the other community, that there are state offices in each community that serve the residents of both communities, that the regional hospital and airport serve both communities, that both Berlin and Barre are members of the Central Vermont Regional Planning Commission and the Central Vermont Waste District, and that the governing bodies of both municipalities are currently meeting to address common issues and problems. In short, there is ample evidence of common interest between the town of Berlin and the city of Barre in support of the creation of the Washington 4 representative district. We find no statutory violation.

D. The Hartland Petition

Petitioners from the towns of Hartland, West Windsor, and Windsor challenge the Legislature’s establishment of two single-member districts, one containing the town of Windsor, and one containing the towns of West Windsor and Hartland. The Legislative Apportionment Board proposed that all three towns remain in a single two-member district, which would have had a *39deviation of -1.6% from the ideal district. The Windsor and West Windsor boards of civil authority expressed their preference for a single two-member district, while the Hartland board wanted two single-member districts. Eventually, the House Government Operations Committee recommended, and the Legislature adopted, a plan creating two districts. The district containing the town of Windsor has a deviation of -1.0%, and the district containing the towns of West Windsor and Hartland has a deviation of 4.2%.

Petitioners argue that the creation of the two districts violates the Equal Protection Clause of the Fourteenth Amendment of the United States Constitution because the new districts do not attain the minimum percentage of deviation practicable. According to petitioners, the State failed to show that the increase in deviation in one of the new districts was justified by a legitimate, rational state interest. We conclude that the establishment of the two districts does not constitute an equal protection violation.

Petitioners rely on Karcher v. Daggett for the proposition that even the most minor variance from equal population must be justified by a legitimate state policy. This position is erroneous. Karcher considered the validity of a congressional redistricting statute, and explicitly stated that the absolute-population-equality requirement applied to apportionment of congressional districts only, not to state legislative districts. 462 U.S. at 732-33 (“we have required that absolute population equality be the paramount objective of apportionment only in the case of congressional districts”). Because we conclude that the 4.2% deviation is de minimis and therefore does not require justification by the State, we need not reach the issue of whether a preference for single-member districts is a legitimate state policy that justifies the slight increase in the deviation of one of the new districts. The deviation complained of here is well within the limits permitted under the United States Constitution.16

*40Petitioners also cite Vermont law in support of their position, although without any supporting argument. The Vermont Constitution requires the Legislature to reapportion the membership of the General Assembly so as “to maintain equality of representation among the respective districts as nearly as it is practicable.” Vt. Const, ch. II, § 73. Further, Vermont statutory law requires the Legislature to form “districts with minimum percentages of deviation from the apportionment standard,” which is the number obtained by dividing the total population of the state by the number of representatives or senators. 17 V.S.A. §§ 1902, 1903(b). These provisions do not require stricter adherence to equal population requirements than that required by federal law. The phrase “as nearly as it is practicable,” which was adopted from federal case law construing the federal equal protection clause,17 contemplates the consideration of other nonnumerical factors. Indeed, Vermont historically has placed great importance on the integrity of individual towns and other nonnumerical criteria, while the existence of state constitutional and statutory provisions requiring equal representation according to population can be traced to federal court decisions construing the federal constitution. In re Senate Bills 177 & 83, 132 Vt. at 287-88, 318 A.2d at 160-61. We find no violation of the Vermont Constitution or Vermont statutory law.

E. The Shrewsbury Petition

Petitioners challenge the placement of the town of Shrews-bury, which is in Rutland County, in a representative district with the towns of Plymouth and Ludlow, which are in Windsor *41County. Petitioners argue that placing Shrewsbury in the same district with Ludlow and Plymouth violates constitutional and statutory criteria that require the Legislature to join towns with common interests. See Vt. Const, ch. II, § 13 (“General Assembly shall seek to maintain geographical compactness and contiguity and to adhere to boundaries of counties and other existing political subdivisions.”); 17 V.S.A. § 1903(b) (insofar as practicable, representative districts should preserve existing political lines, maintain patterns of geography, social interaction, trade, political ties, and common interests, and consist of compact and contiguous territory). Because petitioners have failed to show that there was no rational basis for the creation of the challenged district, we dismiss the petition.

For the most part, the master’s findings are undisputed. From 1982 to 1992, Shrewsbury had been placed in a district with the town of Sherburne. Both towns made it known that the district was unworkable because of inaccessibility between the two towns and because of conflicting positions regarding ski development. Following the 1990 census, the Legislative Apportionment Board placed Shrewsbury in a district with the towns of Tinmouth and Wallingford, and the town of Mount Holly with Plymouth and Ludlow. Rejecting that proposal, the House Government Operations Committee created a plan, eventually adopted by the Legislature, that contained districts in which Shrewsbury and Mount Holly were switched from the Board’s plan. Shrewsbury was joined with Ludlow and Plymouth to form Windsor-Rutland 1 district, while Mount Holly was realigned with Tinmouth, Wallingford, and Mt. Tabor to form Rutland 4 district.

The resulting Windsor-Rutland 1 district is shaped like an inverted “L”, with Shrewsbury and Plymouth across the top from west to east, and Ludlow south of Plymouth. A graded gravel road, which is passable only six months of the year, provides access over the mountain from Shrewsbury to Plymouth. Shrewsbury and Plymouth are largely undeveloped, rural towns that share the Calvin Coolidge State Forest and the Plymouth Wildlife Management Area, which residents of both towns use for recreational purposes. Access from Shrewsbury to Ludlow is via State Route 103, a major east-west artery that passes through Mount Holly. Route 100 runs north from *42Ludlow to Plymouth. Other than the fact that the three towns are in the same transportation district, there was no evidence presented concerning common interests between Shrewsbury and Ludlow.

Upon review of the record, we conclude that petitioners have failed to show that the Legislature violated any constitutional or statutory criteria in creating the Windsor-Rutland 1 district. As noted, the Legislature’s redistricting plans are presumed to be valid, In re Senate Bill 177, 130 Vt. at 361, 294 A.2d at 654; therefore, there is a heavy burden of proof on those challenging them to show that the plans do not meet constitutional or statutory criteria. Davis v. Bandemer, 478 U.S. at 185 (Powell, J., concurring in part and dissenting in part). Petitioners here have not shown- the absence of social, economic, or political ties among the towns in the challenged district. Rather, their evidence is aimed at showing that Mount Holly has more in common with Ludlow and Plymouth than does Shrewsbury. Petitioners focus their attack on the recommendation of the House Government Operations Committee to place Mount Holly, instead of Shrewsbury, with Wallingford, Tinmouth, and Mt. Tabor. Indeed, petitioners argue that placing Mt. Holly rather than Shrewsbury with Wallingford shows that there is no rational basis for the creation of the challenged district.

Petitioners’ arguments are misdirected. The Legislature is not required to accept the Apportionment Board’s plan. See 17 V.S.A. § 1906 (proposal of Board is referred to appropriate committee, after which General Assembly may enact the Board’s plan into law or substitute its own plan). The final plan adopted by the Legislature need only be in conformity with constitutional and statutory criteria; it need not be the best possible plan. See Gaffney v. Cummings, 412 U.S. at 750-51. By the same token, petitioners cannot succeed merely by presenting an alternative plan that appears to meet constitutional and statutory criteria to a greater- degree than the plan chosen by the Legislature. Rather, the burden is on them to show the absence of any rational or legitimate basis for the Legislature’s plan.

Petitioners have failed to meet that burden. The three towns form a geographically compact and contiguous district and are in the same transportation district. Shrewsbury is sep*43arated by mountains from Plymouth but has ready access by highway to Ludlow and through Ludlow to Plymouth. Further, there is a seasonal road directly to Plymouth, and the two communities share a common recreational area. Shrewsbury is not in the same county or school district as Ludlow and Plymouth, but that alone does not indicate there was no rational basis for creation of the district. See In re Senate Bills 177 & 83, 132 Vt. at 289, 318 A.2d at 162 (crossing of county lines does not disqualify plan, given difficulties of reaching an acceptable result statewide); In re Reapportionment Plan for Pennsylvania General Assembly, 442 A.2d at 668 (crossing county lines is unavoidable).

Petitioners, of course, may cast doubt on the validity of a legislative plan by demonstrating that the Legislature failed to adopt an alternative that substantially improves on that plan, see In re Legislative Districting of General Assembly, 193 N.W.2d at 790, but they have also failed to do that. The thrust of petitioners’ argument is that the failure of the House Government Operations Committee to adopt the Board’s proposal aligning Mt. Holly with Ludlow and Plymouth demonstrates the lack of a rational basis for placing Shrewsbury with those towns. We cannot agree. Because both Mt. Holly and Shrews-bury are in Rutland County, both alternatives involve crossing a county line. The degree of compactness is similar with respect to each alternative. As for community of interests, there were various factors for the Committee to consider. Both Shrews-bury and Mt. Holly are connected to Ludlow by the same highway, although Mt. Holly is closer to Ludlow and consequently has more trade with it. Unlike Shrewsbury, Mt. Holly is in the same school district as Ludlow; however, disproportionate tax burdens in the two towns have led to conflicts over attempts to finance a union district school. Further, there is an ongoing dispute between Mt. Holly and Ludlow over the location of their common border, and the towns have conflicting views on the development of Mt. Okemo ski area. Finally, there was testimony that Mt. Holly wanted to remain in the district in which it had spent the last eighteen years and conflicting testimony that some residents of that town preferred the Board’s plan. Although petitioners have challenged the significance of the conflicts between Mt. Holly and Ludlow, we cannot conclude that *44there was no rational basis for the Legislature’s decision to keep Mt. Holly with Wallingford and Tinmouth rather than adopt the Board’s plan. See Ater v. Keisling, 819 P.2d at 301 (redistricting plan upheld where “facts would permit more than one rational conclusion”).

Petitioners argue that the master erred by not allowing a potential candidate to testify that the Speaker of the House admitted to him that Shrewsbury was switched with Mt. Holly to protect an incumbent representative. The State had objected that the testimony was hearsay, and we agree. The testimony was offered to prove the truth of the matter asserted and nothing more. See V.R.E. 801(c), 802.

Further, the proffered testimony was not significant. Petitioners themselves concede that the emphasis in redistricting challenges is not whether there was a discriminatory intent, but rather whether there was a violation of constitutional or statutory criteria. See Gaffney v. Cummings, 412 U.S. at 752 (“It would be idle, we think, to contend that any political consideration taken into account in fashioning a reapportionment plan is sufficient to invalidate it.”); see also Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 266-68 (1977) (discussing ways to determine motive of legislators without placing decisionmaker on the stand). As indicated in our prior discussion, petitioners have failed to show a violation of constitutional or statutory criteria. Consequently, the issue of motive is irrelevant. In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d at 147 (because court had already ruled that redistricting plan did not violate federal or state constitution, any evidence of motives of legislative commission was irrelevant). Indeed, we would be naive if we did not acknowledge that political considerations underlie many redistricting decisions. A search for impermissible political motives would embroil this Court in evaluating the wisdom of redistricting decisions, a function that is inconsistent with our limited role.

Petitioners also allege that creation of the challenged district violates Articles 7 and 8 of Chapter I of the Vermont *45Constitution. See Vt. Const, ch. I, art. 7 (government is instituted for the “common benefit” of the community and not for any particular group within that community); id. ch. I, art. 8 (all freemen, having a common interest with the community, have a right to elect officers). These arguments appear to be an attempt to bootstrap petitioners’ claims into a stricter level of scrutiny than would normally be available in a redistricting petition. The constitutional provisions governing reapportionment are self-contained; there is no indication that additional limits on legislative prerogatives were intended to be applied from other parts of the constitution. Petitioners’ arguments are a restatement of, and are completely dependent on, the claims regarding Chapter II, § 13 and 17 V.S.A. § 1903(b). Petitioners argue that Article 7 applies because the Legislature failed to match Shrewsbury with towns of similar interests, and that Article 8 applies because denying a citizen the right to vote in a district with citizens that share common interests effectively denies that citizen the right to elect a representative. Because these arguments are premised on the lack of common interests within Windsor-Rutland 1, a claim that we have rejected, they must also fail.

Finally, petitioners’argument that the creation of the Windsor-Rutland 1 district denied a certain would-be candidate a fair chance at being elected is without merit. Potential candidates are not guaranteed a favorable district to run in, and citizens of a district do not have a right to vote for a specific candidate. In re 1991 Pennsylvania Legislative Reapportionment Commission, 609 A.2d at 142 (aspirant’s right to run for political office “does not rise to a constitutionally protected level requiring the Legislative Reapportionment Commission to tailor its plan around the residences of political aspirants who seek to challenge a specific incumbent”).

F. The Richford Petition

Petitioners from Richford challenge the Legislature’s redistricting of the Senate, with respect to the Essex-Orleans district, which includes all of the towns within Essex and Orleans counties, and one town each from Franklin, Caledonia, and Lamoille counties. Richford is the lone Franklin County town in the district. Petitioners contend that placing Richford in the *46Essex-Orleans district violates Vermont constitutional and statutory criteria concerning county lines and common interests. See Vt. Const, ch. II, § 18 (in establishing senatorial districts, General Assembly shall seek to adhere to county boundaries); 17 V.S.A. § 1907 (board shall apportion the senatorial seats “among the counties or combinations of counties”); 17 V.S.A. § 1903(b). We dismiss the petition.

The situation here is similar in some, but not all, respects to that in the Montgomery petition challenging its House district. Like Montgomery, Richford is a Franklin County town placed in a district with towns on the other side of the spine of the Green Mountains. Richford is equidistant from the major population centers of Franklin and Orleans counties, St. Albans and Newport, and there is direct access by State Route 105 to both population centers, as well as a more circuitous route through Canada to Newport. Nevertheless, the direct route over the mountains to Newport is difficult during the winter, and, consequently, the ties between Richford and towns in Orleans or Essex counties are minimal. Further, as noted, the weekly and daily newspapers of Franklin and Orleans counties generally do not cover local events outside their own county.

Petitioners first contend, as did the Montgomery petitioners, that it would not offend equal protection requirements if Rich-ford were placed with the Franklin County district, and the other districts were left as they are. This would create an overall deviation of 21.8%, 5.3% greater than the 16.3% deviation in the current plan. It would leave the Essex-Orleans district with 32,228 persons, 14.1% below the Senate apportionment standard. We acknowledge that the difference between the plans is not great, but, as noted, the United States Supreme Court has indicated that an overall deviation of 16.4% approaches tolerable limits. Mahan v. Howell, 410 U.S. at 329. We have rejected an overall deviation of 25.3% for Senate districts. In re Senate Bill 177, 130 Vt. at 370, 294 A.2d at 659. More important, the Legislature’s burden is to maintain “equality of representation ... as nearly as it is practicable.” Vt. Const, ch. II, § 73. An overall deviation of 21.8%, under the circumstances facing the Legislature, was inconsistent with that command.

Petitioners also argue that the Essex-Orleans senatorial district violates constitutional and statutory redistrict*47ing criteria. Specifically, they point out that our constitution requires the Legislature to adhere to boundaries of counties, Vt. Const, ch. II, § 18, and that 17 V.S.A. § 1907 requires that Senate districts be made up of “counties or combinations of counties.” They see in these provisions a prohibition against crossing county lines. We have already rejected this position, In re Senate Bills 177 & 83, 132 Vt. at 286, 318 A.2d at 160, and see no reason to change our holding. County lines must, if necessary, give way to the higher priority of equal representation by population. See id.

Second, they argue that placing Richford with towns in the current district violates the nonnumerical requirements of Chapter II, § 18 (geographical compactness and contiguity) and 17 V.S.A. § 1903(b)(2) (recognition of patterns of geography, social interaction, trade, political ties and common interests). Despite the similarities between this petition and the Montgomery petition, we are persuaded that the differences between them warrant dismissal of this petition. Here, we are dealing with the creation of Senate districts, which, because of the proportion of senators to representatives, are necessarily five times the size of representative districts. Consequently, it will be difficult, if not impossible, to achieve the same level of common interests among individual towns within Senate districts that is attainable among towns within House districts. Two-member senatorial districts contain approximately 37,500 people, which, in most cases, will involve many towns in our rural state. Indeed, the Essex-Orleans senatorial district is comprised of forty-one towns from five different counties.

Considering that the common interests among the residents of towns within Senate districts will invariably be diluted to some extent, the State’s limited showing of common interests within the current Essex-Orleans senatorial district carries more weight. The district is geographically compact and contiguous, but separated by a mountain range. There is road access between Richford and the rest of the district. Also important is the historical treatment of Richford in Senate representation inasmuch as the legislative command is to recognize and maintain existing patterns with respect to nonnumerical factors, including political ties. 17 V.S.A. § 1903(b)(2); see also 17 V.S.A. § 1903(b)(1) (Legislature should preserve existing political sub*48division lines insofar as practicable). Richford has been a member of the Essex-Orleans senatorial district for twenty years, and past and present senators have met with Richford residents and worked with them on a number of issues. We do not mean to imply that there is a presumption of common interests when the district is long standing. We merely point out that the master’s findings recognize some political interaction within the district over the years, and that the Legislature could consider this fact in deciding to keep Richford in a district with towns with which it has long been associated.

Finally, there is evidence of consideration of alternatives in response to Richford’s request. The Senate committee considered moving Caledonia County towns into the Essex-Orleans district to make up for the loss of Richford, but the ripple effects were not manageable.

Based on all of the above considerations, we conclude that petitioners have failed to show that there was no rational or legitimate reason for placing Richford in the Essex-Orleans senatorial district.

IV. Conclusion

Our review of these petitions should reveal at least two important points. The first is that the nonnumerical statutory guidelines contained in 17 V.S.A. § 1903(b) are an implementation and extension of constitutional criteria that seek to achieve effective representation. Accordingly, those policies are important in the redistricting process and must be considered. The second point is that, although we emphasize the requirement that the Legislature consider the nonnumerical statutory criteria, we acknowledge and reiterate that redistricting is a legislative function. The Legislature must have rational and legitimate reasons for the creation of legislative districts based on consideration of all constitutional and statutory criteria. But we are not a superlegislature. If the Legislature’s plans have rational and legitimate bases, we will not disturb them merely because there appear to be better alternative plans or because politics entered into the process.

The Hartland, Springfield, Berlin, Richford, and Shrews-bury petitions are dismissed. Because the inclusion of the town of Montgomery in the Orleans-Franklin 1 district does not ap *49 pear to comply with constitutional and statutory criteria, we return the House plan to the Legislature for reconsideration. The Legislature shall revise and correct the plan so as to place the town of Montgomery in a district that conforms with statutory and constitutional requirements and make any other necessary changes, or provide this Court with information indicating that it is not possible to make such changes while satisfying all constitutional and statutory criteria with respect to the other districts and the plan in its entirety.

concurring and dissenting. I concur in parts I, 111(B) (Springfield), III(C) (Berlin), 111(D) (Hartland), III(E) (Shrewsbury), and III(F) (Richford). I do not concur in part 111(A) (Montgomery) and the standard of review analysis in part II on which that result is based. I would deny the Montgomery petition along with the others. In my view the disposition of the Montgomery petition is an unwarranted interference with the legislative function that will in future years mire this Court in reapportionment petitions for no legitimate purpose.

The entire case put forward by the Town of Montgomery consisted of a short stipulation of the essential facts, five witnesses whose entire examination and cross-examination consists of thirty transcript pages, a few exhibits (most of which are maps) and an affidavit of Professor Frank Bryan of the University of Vermont. The thrust of the town’s ease was that the mountains surrounding Montgomery meant that its economic and social interactions were with Franklin County towns to the east. From that evidence, it built its case that Montgomery should be shifted to the adjoining district containing the Franklin County towns, despite the population disparities such a shift would create. There was absolutely no evidence of the feasibility of reopening the whole districting scheme to satisfy Montgomery. The only mention of an alternative solution was a statement by Professor Bryan that, although he had not “run the data” himself, he would “bet the house and the car” that an alternative could be found that met Montgomery’s demands “and even come closer to the one-person, one-vote criterion.” There was, of course, no specification of such an alternative.

The State’s response was also sparse. In addition to entering into the stipulation of facts, it offered the testimony of Representative Westman, a member of the House Government Opera*50tions Committee, who testified to receiving Montgomery’s arguments and discussing them in committee. He stated that the districts were left undisturbed because of the need to afford equality of representation and because of Montgomery’s ties to the Jay Mountain ski area. From the testimony and the maps, it appears that the ski area is in the Town of Jay and that the southern and eastern access to the ski area is through the Town of Montgomery. Further, it also appears that some Montgomery residents work at the ski area and that Montgomery businesses, primarily restaurants, have customers who come to the area to ski.

Like the Town of Shrewsbury, the petition of which is also addressed today, Montgomery presented one alternative — moving the town to the adjoining Franklin County district. The Court correctly rejects this alternative as inconsistent with Equal Protection requirements. That should end the matter, and the petition should be denied. Without a request, however, and with no understanding of the ripple effects of its action, this Court requires the Legislature to reconsider the whole reapportionment plan in an attempt to satisfy Montgomery, with the Legislature’s only solution being to convince this Court that it is impossible to meet Montgomery’s demand. I cannot agree with this distortion of the proper process and thus dissent.

In my opinion, three major errors in analysis underlie the approach taken by the Court. The first error deals with the standard of review in general, the second with the standard on the Montgomery issues, and the third with the factual analysis. On the first, the Court recognizes that redistricting is primarily a legislative function and that challengers have a heavy burden of proof in showing that a redistricting plan is invalid. It omits, however, that the establishment of districts is done by statute, which is “entitled to the presumptions of justification and regularity accorded regular statutory enactment.” In re Senate Bill 177, 130 Vt. 358, 361, 294 A.2d 653, 654 (1972); see also In re 1983 Legislative Apportionment of House, Senate, & Congressional Districts, 469 A.2d 819, 827 (Me. 1983) (apportionment law entitled to the same presumption of validity as any other legislative enactment).

Attacks on the validity of a statute must be based on “clear and irrefragable evidence that it infringes the paramount law.” *51In re Neglected Child, 129 Vt. 234, 241, 276 A.2d 14, 18 (1971). In equal protection challenges, which are analogous to the challenge the Court considers here, we normally uphold a statute as long as the classification “is not arbitrary and has a reasonable connection with a permissible legislative or administrative purpose.” Veilleux v. Springer, 131 Vt. 33, 39, 300 A.2d 620, 624 (1973). Thus, a reapportionment challenge should prevail only if the redistricting “‘cannot possibly be justified by the exercise of any judgment or discretion.’” Preisler v. Doherty, 284 S.W.2d 427, 431 (Mo. 1955) (quoting State ex rel. Lamb v. Cunningham, 53 N.W. 35, 55 (Wis. 1894)); see also Merriam v. Secretary of Commonwealth, 376 N.E.2d 838, 848 (Mass. 1978) (petitioners must show beyond a reasonable doubt that it is impossible to interpret the reapportionment statute as in harmony with the constitution). Under these standards of review, it is irrelevant whether the Legislature considered all relevant constitutional and statutory criteria with respect to its action. A legislature is judged by its acts, not its processes. See Harrison v. PPG Industries, Inc., 446 U.S. 578, 592 (1980) (in construing a statute, Court does not “in the manner of Sherlock Holmes, pursue the theory of the dog that did not bark”); Townsend v. Yeomans, 301 U.S. 441, 451 (1937) (legislature is presumed to know the needs of the people; whether special inquiries should be made is entirely a matter of legislative discretion); Wall Distributors, Inc. v. City of Newport News, 782 F.2d 1165, 1169 (4th Cir. 1986) (legislature not required to “create an evidentiary record that would pass muster on plenary judicial review of legislation’s necessity and fitness to achieve desired results”); see also Book Review, 66 Yale L.J. 973, 975 (1957) (legislator’s factual assumptions “need be based on no evidence of record”).

In hindsight, part of the problem in this case is that the State offered the evidence of a legislator to defend against the Montgomery challenge. The majority has now treated him, and the House Government Operations Committee on which he sits, like a trial court which must explain and defend its actions. Compare Andreson v. Andreson, 145 Vt. 634, 636, 497 A.2d 371, 373 (1985) (findings must be adequate to explain court’s action and if evidence is inadequate, “the trial court must inquire”). Thus, the legislator was faulted because he “was unable to provide details or support for any connection between the two com*52munities” and because “he never indicated that the Committee considered nonnumerical criteria but was unable to come up with a plan.” It sets a dangerous precedent to determine the validity of a legislative statute based on what evidence its committee heard or considered.

The second error underpinning the Court’s analysis lies in its evaluation of the standards applicable to the Montgomery claim. The relevant constitutional and statutory provisions set forth three types of standards the reapportionment plan must meet. Placed in hierarchical order, these are:

1. Equality of Representation — According to Chapter II, § 73 of the Vermont Constitution, the plan must provide “equality of representation among the respective districts as nearly as it is practicable.” Because of the importance of this criterion, the statute goes further and requires that districts be formed “with minimum percentages of deviation from the apportionment standard.” 17 V.S.A § 1903(b).

2. Nonnumerical Constitutional Standards — According to Chapter II, § 13 of the Constitution, the Legislature “shall seek to maintain geographical compactness and contiguity and to adhere to boundaries of counties and other existing political subdivisions.” These standards are also recognized in the statute with the requirement that districts be formed “consistent with [these] policies insofar as practicable.” 17 V.S.A. § 1903(b)(1), (3).

3. Statutory Standards — Under the requirement that the Legislature form districts “consistent with the following policies insofar as practicable,” the statute includes “recognition and maintenance of patterns of geography, social interaction, trade, political ties and common interests.” 17 V.S.A. § 1903(b)(2).

That these requirements should be viewed in hierarchical fashion is clearly shown by their placement. Cf. In re Senate Bill 177, 130 Vt. 365, 371, 294 A.2d 657, 660 (1972) (preeminence of constitutional requirements). Further, the provisos attached to these standards recognize that it may be impossible to meet all of them in any given case, and that a balancing of factors is inevitable. I am not sure I have any disagreement with the majority on these general points. I have differences in the specifics.

*53The majority concludes that “petitioners have shown that none of the nonnumerical statutory or constitutional criteria were adhered to with regard to the Town of Montgomery.” It reaches that conclusion, I infer, by reasoning that the “compactness and contiguity” requirement really means community of interest or creation of “effective representation,” that as a result there are two nonnumerical requirements — community of interest/effective representation and adherence to political boundaries — and that the placement of Montgomery with the Orleans county towns meets neither of these standards.* The reasoning is clearly flawed.

I can agree that one of the purposes of the compactness and contiguity requirements is to create a commonality of interests in legislative districts. See, e:g., Karcher v. Daggett, 462 U.S. 725, 756 (1983) (Stevens, J., concurring) (“To some extent, geographical compactness serves independent values; it facilitates political organization, electoral campaigning, and constituent representation.”). The majority has, however, confused a purpose with the requirement. Compactness and contiguity are clearly geographic requirements. See Carpenter v. Hammond, 667 P.2d 1204, 1218-19 (Alaska 1983) (Matthews, J., concurring); Acker v. Love, 496 P.2d 75, 76 (Colo. 1972); Schrage v. State Board of Elections, 430 N.E.2d 483, 486 (Ill. 1981). That is clearly so in Vermont because the constitution requires “geographical compactness and contiguity” (emphasis supplied). Vermont Const, ch. II, § 13. There is nothing in the requirement that suggests a district is not compact or contiguous because it has a mountain in it.

If we are to examine the purposes of the compactness and contiguity requirement, it is important to recognize that the dominant purpose is to prevent gerrymandering. See, e.g., Holmes v. Farmer, 475 A.2d 976, 986 (R.I. 1984); see generally J. Schwartzberg, Reapporbionment, Gerrymanders, and the Notion of “Compactness,” 50 Minn. L. Rev. 443 (1966). To the extent compactness has a nongeographic component, it lies in the prevention of “districts solely for political considerations, *54without reference to other policies.” Holmes, 475 A.2d at 986. The Legislature recognized that compactness and contiguity is different from community of interest because it set them out separately. See 17 V.S.A. § 1903(b)(2), (3).

The California case on which the majority relies to equate compactness and contiguity to community of interest, Wilson v. Eu, 823 P.2d 545, 553, 4 Cal. Rptr. 2d 379, 387 (1992), demonstrates the danger in relying on legislative apportionment cases from other states without careful examination. California has no constitutional requirement that legislative districts be compact. See Cal. Const, art. 21, § 1. Further, the reapportionment process in California had broken down to the point where redistricting had to be done by the courts. Thus, the phrases quoted by the majority refer to judicial standards created for judicial reapportionment and not to constitutional standards to judge the validity of a legislatively-created districting scheme. Any legislature would be wise to consider communities of interest in drawing districts. It is quite a different thing to say that the failure to do so in a particular case makes a district unconstitutional.

Each town in the challenged Orleans-Franklin District is contiguous with at least one other town. Montgomery is contiguous with the towns of Westfield and Lowell. The district is as compact as any other multi-town district in the state. No one has challenged the district as lacking in compactness or contiguity. There is no serious claim that it violates either requirement.

Also significant, in view of the majority’s reliance on the purpose of contiguity and compactness requirements, is the lack of any evidence even hinting that this district was gerrymandered for political purposes, and the Town of Montgomery does not so claim. Thus, the district meets both the letter of the compactness and contiguity requirements and their primary purpose.

The district does cross the county boundary, and in that sense is inconsistent with the “goal” of chapter II, § 13. It shares that characteristic with twenty other multi-county districts in the state. Except as a passing note, little is made of the breach of the county line by either the majority or the petitioners. County lines are of limited significance to house districts because of the very limited county government in Vermont. Where the county line follows a geographical boundary, the breach of the county *55line is significant because of the geographical boundary, not necessarily because of the county line.

More significant are town lines, and this district crosses none of these. It is impossible in a rural state with a large number of towns to follow town lines without crossing county lines. Thus, the district here is consistent with the important boundary requirement of § 13.

Instead of the total violation of nonnumerical requirements claimed by the majority, the reality is that the district in question complies fully with the significant, nonnumerical constitutional requirements. Thus, the real complaint is that the Legislature failed to abide by a statutory requirement. As discussed below, I find the evidence of that failure shallow and incomplete. For purposes of our approach to review, it is important to state how vague and tentative this statutory requirement is. It requires that districts be formed “consistent with the following policies insofar as practicable” and details one of those policies as “recognition and maintenance of patterns of geography, social interaction, trade, political ties and common interests.” 17 V.S.A. § 1903(b)(2). Given the breadth and imprecision of the requirement, and its placement at the bottom of the hierarchy of requirements, we should be saying that only the clearest and most extreme breaches of the mandate are cognizable by this Court in reviewing a reapportionment plan. See Fonfara v. Reapportionment Commission, 610 A.2d 153, 163 (Conn. 1992). Our failure to do so means that this Court will make the choices that determine district boundaries rather than leaving those choices to the appropriate body, the Legislature.

The majority has avoided confronting the most difficult problem caused by its analysis. If the only real deviation from applicable standards is in compliance with 17 V.S.A. § 1903(b)(2), as I believe it to be, then why cannot the Legislature react by simply repealing the statute? We should be very careful to adopt a mandate subject to that kind of evasion. We have not been careful here.

Finally, I think the majority has accepted a shallow and incomplete notion of community of interest that ignores the reality of the tasks facing the legislators who represent these districts. This opinion is issued at the start of a legislative ses*56sion in which the major concerns are the level of state taxation and the nature of those taxes, the level of state- spending and the areas in which the spending occurs, creating jobs and income growth to counteract the effect of the recession, and regulatory reform. There is absolutely nothing in the sparse evidence produced by Montgomery to show that the residents of that town have a common interest in how these issues are resolved with the residents of the Franklin County towns in the adjoining district.

One does not have to consider the legislative issues in depth to realize that a significant possibility exists that the real common interests here are between the “mountain” towns. -That is, if a significant portion of jobs in Montgomery are dependent on tourists who come to ski at Jay Peak, its citizens may see economic development as promotion of tourism, through state spending and tax policy, as well as regulatory reform to make such development easier. This interest is shared with residents of towns like Jay but not necessarily with the towns of Franklin County, which depend on an agricultural economy.

Although this evaluation might be considered speculative, it is useful to note a significant similarity in the facts of the Shrewsbury appeal. Shrewsbury had been joined in a district with the Town of Sherburne, with which it shared the kind of social and economic interaction demonstrated in the evidence put forward by Montgomery. That joinder failed, however, because the towns disagreed on ski area development and related environmental issues. Social, political and educational ties are important, but they do not mean that citizens with these ties have a common perspective and interest in the policy issues a legislature must resolve. For example, the difficulty of transportation and communication over the mountain raises a common legislative issue for Montgomery and its abutting Orleans County towns. The Montgomery evidence showed little about commonality of interests in the legislative issues of the day. It is at best incomplete.

I find it particularly ironic that the majority ignores the evidence that represents the most significant community of interest factor. Although the majority recognizes that Representative Westman’s testimony indicated a reliance, in part, on the Jay Peak connection, it goes on to say that the only reason for *57the district formation was avoidance of “an equal protection problem.” That is a misstatement of the evidence. Significantly, the Jay Peak connection may be the most important element in the “patterns of geography, social interaction, trade, political ties and common interests” present. See 17 V.S.A. § 1903(b)(2). Certainly, the Legislature is entitled to draw that conclusion without being second-guessed by the judiciary.

These three major points of disagreement with the reasoning of the Court lead to my disagreement with its conclusion both on process and result. Ultimately, the Court derives from its analysis a burden-shifting rule under which the showing of lack of community of interest by petitioner shifts to the Legislature the burden of showing “that an alternative plan satisfying the various constitutional and statutory criteria could not be produced.” Failing this burden, the Legislature is directed to place the Town of Montgomery in a new district “that conforms with statutory and constitutional requirements and make any other necessary changes” or to show that such a change is “not possible.” In this process, petitioners are not required to demonstrate that there exists at least one alternative districting plan that meets the constitutional and statutory mandates and places Montgomery in a district with Franklin County towns.

I find the burden-shifting rule totally inconsistent with our limited standard of review. I would follow the recent and thoroughly-reasoned opinion of the Connecticut Supreme Court in Fonfara v. Reapportionment Commission, 610 A.2d at 159, that a mere showing of a violation of a nonnumerical requirement — in that case, a state constitutional prohibition on crossing town lines — does not shift the burden of proof to the apportionment body to prove the plan is valid. If anything, this case is easier than Fonfara because the requirement here is statutory rather than constitutional, less specific, and petitioners have presented no alternative plans.

I find no parallels in any of our standards of judicial review of legislative action to the burden-shifting and “remand” procedure created by the majority. The “heavy burden of proof” imposed by the majority on petitioners miraculously disappears. Nor are there any precedents for such a “remand” to the Legislature in the opinions from other states. Whenever a “remand” has been used, it has been to a redistricting commission, *58a type of administrative agency over which the court has a form of judicial review. See In re Colorado General Assembly, 828 P.2d 185, 195-96 (Colo. 1992) (under a specific constitutional mandate that “communities of interest. . . shall be preserved” and a specific constitutional procedure that requires a reapportionment commission to submit a plan to the Colorado Supreme Court “for review,” failure of commission to explain why it combined parts of one county with other counties across the continental divide required a remand for a factual showing that less drastic alternatives were unavailable); Davenport v. Apportionment Commission, 304 A.2d 736, 746-47 (N. J. Super. 1973) (redistricting plan remanded to reapportionment commission because recent decision of United States Supreme Court suggested reapportionment could be done while respecting county lines as required by New Jersey Constitution; commission directed to report on whether compliance with state constitution “is possible”). The court’s relationship to an administrative agency is far different from its relationship with a legislature.

Nor have any other courts relieved petitioners from the burden of showing that at least one alternative valid plan exists. The cases cited by the majority hold that such a showing does not mean the petitioner will automatically prevail. The one court to analyze the issue has held that a showing of a valid alternative is required in a situation similar to that present here. In In re 1988 Legislative Apportionment of House, Senate, & Congressional Districts, 469 A.2d at 831, petitioners attacked a house district as not meeting the compact and contiguous requirement of the Maine Constitution. The court ruled the attack unsuccessful because changes in the district would produce a “ripple effect” elsewhere and “a challenger at the very least must show that . . . [the] district . . . could be substantially improved without creating constitutional violations elsewhere in the state.” Id. This is exactly the burden the petitioners failed to meet here.

The burden allocation is directly contrary to the rules under which we normally allocate burdens of proof. As discussed above, the statutory mandate of community of interest must be met “insofar as practicable.” 17 V.S.A. § 1903(b). The majority has stated that the burden of proof of noncompliance with constitutional or statutory mandates lies on petitioners and that it *59is a heavy burden. See majority opinion at 14, 624 A.2d at 326. Despite this stated allocation, it relieves petitioners of the only important burden of proof involved in this case — the burden of showing what is practicable. The fact is that petitioners have never demonstrated a noncompliance with § 1903(b)(2) and are not being required to do so.

The only rationale given for the burden-shifting rule is that “basic notions of fairness” require the burden be placed on the State because the Legislature can better devise alternative plans and because other towns will be affected. The effect on other towns is exactly the point. They should have an opportunity in this Court to respond to any proposed change in their circumstances rather than being blind-sided by an unpredictable direction to the Legislature. Nor do I believe that the burden to develop an alternative plan is heavy in this small state. Indeed, alternatives have routinely been presented by petitioners in other cases. See, e.g., Hellar v. Cenarrusa, 664 P.2d 765, 768 (Idaho 1983); In re Legislative Districting of General Assembly, 193 N.W.2d 784, 790 (Iowa 1972); Clements v. Valles, 620 S.W.2d 112, 115 (Tex. 1981). In Davenport v. Apportionment Commission, the New Jersey Superior Court was struck that “one of the parties, on very short notice and without the aid of computer services, produced a plan which, at least on its face, represents a far greater compliance with the requirement of compactness while still complying substantially with the equal population imperative.” 304 A.2d at 743-44. I cannot believe what can be done “on very short notice” in New Jersey is an unacceptable burden here.

Finally, I can find no precedent in which a legislative apportionment scheme was upset under the type of broad, statutory standard involved here and where such weak evidence of noncompliance with a determinative standard was presented. It is inescapable that, when we begin to judge legislative action by whether it respects a proper community of interest or creates effective representation, we have taken on ourselves the reapportionment policy decisions even while representing that we are leaving the legislative role intact. I think we will come to regret the day we took on this superlegislative responsibility.

The towns of Vermont who are not presently before this Court will be surprised to learn that the redistricting plan that *60they have just accepted has now been reopened to cope with the ripple effect resulting from the “repair” of the placement of the Town of Montgomery. As with the Senate District placement of Richford, it may well turn out that the possible alternatives are far less desirable than the original plan. Unfortunately, the Court’s mandate overrides legislative judgments of desirability. The analysis and ruling that reaches that result is unwise, inadequately supported and inappropriately intrudes on a legislative function. I dissent.

I am authorized to state that Justice Morse joins in this dissent.