Joseph J. REGENSBURGER; Jeremy Neff, et al., Plaintiffs-Appellees, v. CITY OF BOWLING GREEN, OHIO, et al., Defendants-Appellants, 278 F.3d 588

COLE, J., delivered the opinion and judgment of the court, BATCHELDER, J., concurring in the judgment of the court but dissenting as to the reasoning employed in

Part III.F. (pp. 598). JOHN R. GIBSON, J., joins in J. BATCHELDER’S separate writing as to

Part III.F. which becomes the opinion of the court. OPINION COLE, Circuit Judge. Plaintiffs-Appellees, residents of Bowling Green, Ohio, including students at Bowling Green State University (“BGSU”) (all collectively referred to herein as “Plaintiffs”), initiated the present action against Defendants-Appellants, the City of Bowling Green, Ohio, and various city officials (collectively referred to herein as “the City”), alleging that the City’s voting apportionment scheme violated their rights under the Fourteenth Amendment’s Equal Protection Clause.

After a bench trial, a magistrate judge found that, the City’s apportionment plan deviates from absolute population equality by at least 66.85%; that the City had set forth a rational state policy for the deviation; that the plan advanced the articulated policy; and that, the articulated policy notwithstanding, the population deviation exceeded constitutional limits. The magistrate judge ordered the City to develop and submit to the court for its review and approval a constitutionally permissible reapportionment plan.

The district court affirmed the judgment of the magistrate judge in its entirety, and the City now appeals, assigning error to the magistrate judge’s determinations that:

(1) Bowling Green is subject to the requirements of Ohio Revised Code § 731.06; (2) the population deviation among the City’s wards was not mitigated sufficiently by the City’s provision of at-large representation; (3) the City must develop a reapportionment scheme, without guidance from the court, as to the appropriate population base to be used; (4) the 1995 statistics generated by the United States Department of Housing and Urban Development were “unsupported by sufficient facts,” thus requiring the parties to rely upon the 1990 Federal Census population figures; and, (5) although the City’s policies upon which the present apportionment plan is based are rational, the legislative plan is nevertheless unconstitutional.

For the reasons that follow, we AFFIRM the order of the district court in its treatment of the five assignments of error. We REVERSE the judgment of the district court to retain jurisdiction over this case for the purpose of reviewing the constitutionality of the City’s proposed reapportionment scheme. I. BACKGROUND Bowling Green, Ohio, is centered at the crossroads of U.S. Route 6 and U.S. Route 25. These bisecting highways establish a natural division of the city into four geographic quadrants, which, since the city’s inception in 1866, have constituted four legislative wards.

The Bowling Green City Charter (“Charter”) provides that Bowling Green’s City Council shall consist of three members selected “at large” based upon the total number of votes received, and one council member selected from each of the city’s four wards. This apportionment scheme was incorporated into the Charter on October 31,1972, when the municipality became a city.

Although Routes 6 and 25 form a natural geographic division of the city, they do not divide the city into quadrants containing proportional numbers of the city’s population, due in part to the student population of BGSU, which is situated in Precinct 1-C of Ward 1. The City maintains that it, like other “college towns” with significant populations of students, properly substitutes the number of registered voters in the specific “campus” precincts for the Census-based populations when reviewing district representation, effectively counting only registered voters in areas of BGSU students.

Of the 12,674 residents of Ward 1, 7311 of them are registered to vote. The City argues that when adjusted for numbers of registered voters, the population figures reveal that Ward 1 contains 31.9% of the total population, Ward 2 contains 27.5%, Ward 3 contains 15.5%, and Ward 4 contains 25.1%, resulting in a constitutionally permissible population deviation from ideal district size. In early 1995, Plaintiffs challenged the apportionment scheme as a violation 'of the Fourteenth Amendment’s Equal Protection Clause and sought declaratory and injunctive relief from the City.

On September 15, 1995, a magistrate judge granted Plaintiffs’ motion for partial

summary judgment, finding that Plaintiffs had established a prima facie case of discrimination, and leaving for trial only resolution of the question whether the City’s stated reasons for maintaining the ward boundaries provided a rational and constitutionally permissible justification for the population deviations. At an October 17, 1995, bench trial before the magistrate judge, Plaintiffs argued that the City’s plan violated the “maximum population deviation” standard of 10% among legislative districts as required by Voinovich v.

Quilter, 507 U.S. 146, 161-62, 113 S.Ct. 1149, 122 L.Ed.2d 500 (1993) (noting that a prima facie case of discrimination is established upon a showing that an apportionment plan has a maximum total deviation from ideal district size in excess of 10%). After testimony by various city, officials, the magistrate judge found that Plaintiffs had alleged a prima facie case of discrimination, irrespective of whether one used the Census-based population figures or the registered-voter.radjusted population numbers for student precincts in Ward 1.

Although the magistrate judge also found that the state policy justifying the population adjustments advanced by the City was rational, she nevertheless concluded that the resulting population disparity rendered the City’s present apportionment scheme constitutionally defective. The court ordered the City to develop a reapportionment plan and retained jurisdiction to review and approve the new apportionment scheme. We granted the City’s motion for leave to appeal. This appeal follows. II. STANDARD OF REVIEW We review de novo the district court’s

interpretation of the relevant law, but review its factual findings for clear error only. See Thornburg v. Gingles, 478 U.S. 30, 79, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986); see also Cousin v. McWherter, 46 F.3d 568, 574 (6th Cir.1995) (“A district court’s factual findings regarding

Section 2 violations and the determination, of whether vote dilution has occurred are ordinarily reviewed for clear error.)” (citing Fed. R.Giv.P. 52(a)). III. DISCUSSION A. Charter City Bowling Green, a charter city subject to

Article XVIII,

Section 7, of the Ohio Constitution, is governed by a charter that provides in pertinent part that: [t]he legislative power of the city shall be vested in a city council which shall be a continuing body consisting of 3 members at large and one member from each of such even number of wards which shall be established by council by ordinance enacted not later than January 15 of the year in which the first election for members of council from wards as established by such ordinance is to be held.

Ohio law provides that “the legislative authority of a city shall subdivide the city into wards, equal in number to the members of the legislative authority to be elected from wards” and delineates the manner in which a city is to carry out this directive. Ohio Rev.Code Ann. § 731.06 (Anderson 2000).

Relying on the language of § 731.06, the magistrate judge found that “[a] city council has the continuing authority and duty to redivide the city into proper wards pursuant to law when an increase in population occurs.” The City seizes upon this statement, arguing that the magistrate judge erroneously concluded that a city is required to establish a ward system and that the City’s hybrid at-large/ward system is, therefore, impermissible as a matter of law. Citing Pesek v.

City of Brunswick, 794 F.Supp. 768 (N.D.Ohio 1992), for the proposition that a charter city subject to “home rule” may enact a provision that is in conflict with state law, the City argues that, because the Charter provision allowing a hybrid system is in direct conflict with § 731.06’s requirement that a city establish a ward system of representation, the magistrate judge erred by requiring the City to adopt a ward system. The City’s argument is without merit.

In the first instance, the magistrate judge did not make a finding that the at-large system was prohibited by § 731.06 nor did she conclude that the City could not adopt an at-large system. Rather, she found that, even when taking at-large representation into account, the deviation from ideal did not pass constitutional muster. Second, and more important, Pesek does not create legal authority for the City to avoid the reach of § 731.06.

While it is true that “[t]he general rule is that in matters of local self-government, if there is a conflict between a charter provision and a statute, the charter provision prevails,” the City has shown no conflict between its Charter and the statute. State ex rel. Bardo v. City of Lyndhurst, 37 Ohio St.3d 106, 524 N.E.2d 447, 450 (Ohio 1988). The Charter requires at-large and direct ward representation; § 731.06 merely directs the legislative authority of a city that employs a ward system of representation to subdivide the city into wards.

Neither the organization provisions of § 731.06 nor the magistrate judge’s order requires any particular form of apportionment scheme. Thus, there is no conflict that would require this Court to determine which provision must prevail in this case. Even assuming there were in fact a conflict, “the rule of charter supremacy applies only where the conflict appears by the express terms of the charter and not by mere inference.” Lyndhurst, 524 N.E.2d at 450 (emphasis added). The City’s argument necessarily relies upon inference and must be rejected. B.

At-Large Elected Officials It is well-settled that when examining legislative redistricting, a reviewing court must take into account the presence of at-large representation. See Board of Estimate of City of New York v. Morris, 489 U.S. 688, 109 S.Ct. 1438, 103 L.Ed.2d 717 (1989). The City’s contention that the magistrate judge erred in failing to give due regard to the City’s at-large and direct representation system finds no support in the record. It is clear that the magistrate judge did’ consider the existence of the City’s at-large ward system in examining the overall voting scheme.

The magistrate judge made an express finding that “a deviation as high as 66.85% is not constitutionally permissible even though that deviation is offset by the existence of three at-large council positions.” Even if this Court were inclined to look beyond the plain language of the magistrate judge’s opinion, we would nevertheless find that the deviation in this case is unconstitutional.

In Morris, the Supreme Court held that, in determining the deviation among districts, the relevant inquiry is “whether ‘the vote of any citizen is approximately equal in weight to that of any other citizen,’ ... the aim being to provide ‘fair and effective representation for all citizens.’ ” 489 U.S. at 700, 109 S.Ct. 1433 (quoting Reynolds v. Sims, 377 U.S. 533, 579, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964)).

Although the Morris Court concluded that the lower courts erred in failing to determine the deviation from ideal with the at-large representation factored into the analysis, it refused to find reversible error, because the city had conceded a population deviation of 78%. See id. at 701-02, 109 S.Ct. 1433. In the instant case, the magistrate judge found “a deviation as high as 66.85%.” The City does not assert that the magistrate judge failed to calculate the deviation from ideal, nor does it assert that the deviations found by the magistrate judge were within per se constitutional limits. It cannot.

Even assuming the magistrate judge failed to consider the presence of at-large representation in the City, the United States Supreme Court has suggested that a deviation as high as that at issue in this case might be per se unconstitutional. Cf. id. at 702, 109 S.Ct. 1433 (“We note that no case of ours hás indicated that a deviation of some 78% could ever be justified.”).

Given the evidence that the magistrate judge did consider the at-large system in calculating the population deviation from ideal, and given the City’s failure to show that this figure is inaccurate, we conclude that the magistrate judge’s findings were not clearly erroneous. C. HUD Estimates The City contends that the district court improperly rejected use of the 1995 United States Department of Housing and Urban Development (“HUD”) population estimates as “unsupported by sufficient facts.” Use of the HUD estimates, the City argues, would have resulted in a constitutionally permissible deviation of 16.5%.

The City’s argument is without merit. After hearing the testimony of the City Administrator, which called into question the source and accuracy of the HUD estimates, the magistrate judge concluded that use of the 1990 Census figures would provide a more reliable population base for determination of the deviation from ideal. We do not find that the magistrate judge’s determination was clearly erroneous, particularly in light of the fact that, unlike that of the HUD estimates, there was no question as to the accuracy of the 1990 Census figures, to which even the City stipulated. D.

Population Base Although the magistrate judge found that the City’s system of counting Ward 1 residents was rational — a point that neither party disputes — she concluded that, “[gjiven that either method of calculating population deviation exceeds constitutional limits, the court determines that it is unnecessary to reach the issue of whether the reduction of population by registered voters in only one ward is an acceptable method of calculating population.” More important, in striking down the City’s apportionment scheme, the magistrate judge properly refused to state whether, in developing a reapportionment scheme, the City would be required to rely upon the Census figures or the registered-voter-adjusted figures.

Such a question was not properly before the district court, and any statement as to the appropriate population base on which to base a reapportionment scheme would have amounted to nothing more than an impermissible advisory opinion. See North Carolina v. Rice, 404 U.S. 244, 246, 92 S.Ct. 402, 30 L.Ed.2d 413 (1971) (reiterating that federal courts have no power to issue advisory opinions); Church of Scientology of Hawaii v. United States, 485 F.2d 313, 314 (9th Cir.1973) (“[T]he court does not render advisory opinions or decide abstract propositions.”); Minnesota v.

United States Steel Corp., 438 F.2d 1380, 1384 (8th Cir.1971) (“The legal questions should not be considered in the abstract.”). E. Legislative Plan It is well-established that a plaintiff may establish a prima facie case of dilution by demonstrating a population deviation in excess of 10%. See Quilter, 507 U.S. at 161, 113 S.Ct. 1149.

Such a showing shifts to the state or municipality the burden of justifying the deviation, which requires it: (1) to articulate a “rational state policy” that may justify the deviation; (2) to explain how the apportionment plan “may reasonably be said to advance” the rational state policy; and (3) to demonstrate that the resulting deviation does not “exceed constitutional limits.” Mahan v. Howell, 410 U.S. 315, 328, 93 S.Ct. 979, 35 L.Ed.2d 320 (1973); accord Quilter, 507 U.S. at 161-62, 113 S.Ct. 1149.

While it is true that the magistrate judge found that the City’s proffered justification — maintenance of “the efficient and economic delivery of city services in Bowling Green” — was a rational state policy, and that the City’s apportionment plan advanced that policy, she nevertheless concluded that a population deviation as high as 66.85% exceeded constitutional limits, irrespective of the underlying policy reasons for that deviation.

The City maintains that the magistrate judge erred by concluding that the City’s legislative plan was unconstitutional after first finding that the municipal policies underlying the scheme were rational and justified. The City does not dispute the accuracy of the magistrate judge’s finding of a population deviation of 66.85%.

Rather, it seems to argue that a policy that is shown to be rational and advanced by a municipality’s apportionment scheme must necessarily be constitutional, regardless of the attendant population deviation, and that a reviewing court may not conclude, as the magistrate judge did in this case, that a population deviation of 66.85% is per se unconstitutional. The City fails to recognize that the first two prongs of the Mahan test do not exist to substantiate the third.

While it is true that a state has a legitimate interest in maintaining boundaries, see Howell, 410 U.S. at 325-27, 93 S.Ct. 979, such an interest in no way justifies every deviation from ideal. Apportionment schemes that are motivated by rational and otherwise justifiable polices may nevertheless result in population deviations that exceed constitutional limits. The City’s reliance on Brown v. Thomson, 462 U.S. 835, 103 S.Ct. 2690, 77 L.Ed.2d 214 (1983), to support its argument is misplaced. Brown never dealt with the question of constitutionally permissible population deviations.

Rather, it addressed whether Wyoming’s plan for preserving boundaries justified the additional deviations from population equality resulting from a particular county’s representation, focusing on the appellants’ challenge to representation of that particular county and not on the deviations from ideal. See id. at 846, 103 S.Ct. 2690. The Brown Court stated: Here we are not required to decide whether Wyoming’s nondiscriminatory adherence to county boundaries justifies the population deviations that exist throughout Wyoming’s representative districts.

Appellants deliberately have limited their challenge to the alleged dilution of their voting power resulting from the one representative given to Niobrara County. The issue therefore is not whether a 16% average deviation and an 89% maximum deviation, considering the state apportionment plan as a whole, are constitutionally permissible. Rather, the issue is whether Wyoming’s policy of preserving county boundaries justifies the additional deviations from population equality resulting from the provision of representation to Niobrara County. Id. at 846, 103 S.Ct. 2690.

In the present action, the magistrate judge properly limited Brown to its facts and concluded that the deviation at issue in the instant case was not permissible even in light of the arguably mitigative effect of at-large representation. The United States Supreme Court has consistently found that deviations from ideal— while not dispositive — are important in determining whether the challenged legislative plan complies with the Equal Protection Clause. See, e.g., Chapman v.

Meier, 420 U.S. 1, 22, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975) (internal citation and quotation marks omitted) (“WTiile (m)athematical exactness or precision is not required, there must be substantial compliance with the goal of population equality.”). The City’s failure to cite even one case in which a reviewing court has upheld a population deviation as high as 66.85% is especially revealing in light of the overwhelming authority to the contrary. See, e.g., Morris, 489 U.S. at 702, 109 S.Ct. 1433 (suggesting that a 78% deviation could never be justified); Connor v.

Finch, 431 U.S. 407, 418, 97 S.Ct. 1828, 52 L.Ed.2d 465 (1977) (“The maximum population deviations of 16.5% in the Senate districts and 19.3% in the House districts can hardly be characterized as de minimis; they substantially exceed the ‘under-10%’ deviations the Court has previously considered to be of prima facie constitutional validity only in the context of legislatively enacted apportion-ments.”); Chapman, 420 U.S. at 22, 95 S.Ct. 751 (“We believe that a population deviation of that magnitude [20.14%] in a court-ordered plan is constitutionally impermissible in the absence of significant state policies or other acceptable considerations that require adoption of a plan with so great a variance.”); Howell, 410 U.S. at 328, 93 S.Ct. 979 (emphasis added) (noting that although “the 16-odd percent maximum deviation that the District Court found to exist in the legislative plan for the reapportionment of the House .... may well approach tolerable limits, we do not believe it exceeds them.”); Kilgarlin v.

Hill, 386 U.S. 120, 122, 87 S.Ct. 820, 17 L.Ed.2d 771 (1967) (“[I]t is quite' clear that unless satisfactorily justified by the court or by the evidence of record, population variances of the size and significance evident here [26.48%] are sufficient to invalidate an apportionment plan.”); Swann v.

Adams, 385 U.S. 440, 442, 87 S.Ct. 569, 17 L.Ed.2d 501 (1967) (“De minimis deviations are unavoidable, but variations of 30% among senate districts and 40% among house districts can hardly be deemed de minimis and none of our cases suggests that differences of this magnitude will be approved without a satisfactory explanation grounded on acceptable state policy”). Accordingly, we find that the magistrate judge properly concluded that the population deviation exceeded constitutional limits. F.

Jurisdiction Upon determining that the City’s voting apportionment scheme was unconstitutional, the magistrate judge struck down the scheme and ordered the City to develop and submit to the court for its review and approval a constitutionally permissible reapportionment plan. With respect to jurisdiction, the magistrate judge’s memorandum and order stated: In view of the foregoing, Defendants are hereby ORDERED to develop a plan for reapportionment within six months from the date of this judgment and to submit such plan to the Court. This Court retains jurisdiction for review and approval of such plan.

The City did not raise the magistrate judge’s decision to retain jurisdiction in its appeal to the district judge. The district judge affirmed the magistrate judge’s opinion in toto. Likewise, the issue of retention of jurisdiction was not raised in this appeal, and was not addressed by either party in either the initial or supplemental briefs, or at oral argument.

Although neither party addresses this issue, “every federal appellate court has a special obligation to ‘satisfy itself not only of its own jurisdiction, but also [of] that of the lower courts in a cause under review,’ even though the parties are prepared to concede it.” Arizonans for Official English v. Arizona, 520 U.S. 43, 73, 117 S.Ct. 1055, 137 L.Ed.2d 170 (1997) (quoting Mitchell v. Maurer, 293 U.S. 237, 244, 55 S.Ct. 162, 79 L.Ed. 338 (1934)).

In considering whether the lower court’s retention of jurisdiction over the City plan is “necessary or practicable to achieve compliance” with constitutional mandates, we use an abuse of discretion standard. See Missouri v. Jenkins, 515 U.S. 70, 89, 115 S.Ct. 2038, 132 L.Ed.2d 63 (1995) (quoting Freeman v. Pitts, 503 U.S. 467, 491, 112 S.Ct. 1430, 118 L.Ed.2d 108 (1992)); United States v. W.T. Grant Co., 345 U.S. 629, 633, 73 S.Ct. 894, 97 L.Ed. 1303 (1953).

Under this standard, the magistrate judge’s decision to retain jurisdiction will be disturbed if it “relied upon clearly erroneous findings of fact, improperly applied the governing law, or used an erroneous legal standard.” Blue Cross & Blue Shield Mut. of Ohio v. Blue Cross & Blue Shield Ass'n 110 F.3d 318, 322 (6th Cir.1997). In this case, the magistrate judge retained jurisdiction to review and approve the City’s new plan for reapportionment without any factual or legal justification.

We find that this failure to include any stated factual or legal basis for the need to retain jurisdiction is equivalent to “improperly applying] the governing law, or us[ing] an erroneous legal standard,” and constitutes an abuse of discretion. Id. We therefore hold that the magistrate judge improperly retained jurisdiction over this case for the limited purpose of reviewing the constitutionality of the City’s proposed reapportionment scheme, and reverse the judgment of the district court on this point. IV. CONCLUSION Accordingly, we .AFFIRM in part and REVERSE in part the judgment of the district court. .

The 1990 Federal Census revealed that the population of each of the four wards was as follows: Ward 1 population 12,674 (44.98% of total population) Ward 2 population 6,287 (22.31% of total population) Ward 3 population 3,499 (12.42% of total population) Ward 4 population 5,716 (20.29% of total population) . Commonly referred to as the "Home Rule” provision,

Article XVIII,

Section 7, provides that "[a]ny municipality may frame and adopt or amend a charter for its government and may, subject to the provisions of

section 3 of this article, exercise thereunder all powers of local self-government.” Ohio Const, art. XVIII, § 7. . The City seems to argue that the absence from § 731.06 of any reference to an at-large system of representation is evidence that the Ohio legislature favored exclusive use of a ward system. Such an

interpretation, if accepted, would still create no conflict between it and the Charter, as the Charter is silent on the question of reapportionment. Furthermore, as a practical matter, an at-large system of representation would never require reapportionment as such; any statutory provision on the question of reapportionment would therefore be both unnecessary and meaningless. .

Even assuming that the question of the appropriate population base for purposes of reapportionment were properly before this Court, we would reject the City's argument that it should be permitted to rely upon the Census-based population numbers for all parts of the city except Precinct 1-C, where it would substitute the number of registered voters. Such a proposition finds no support in the law, and would violate the firmly established "one person-one vote” principle in the most literal sense.