John W. Cross v. Lloyd Baxter

604 F.2d 875, 1979 U.S. App. LEXIS 11168

• Court: Court of Appeals for the Fifth Circuit
• Decided: October 16, 1979
• Docket: 77-3286
• Status: Published

Opinion

GODBOLD, Circuit Judge:

Members of the Moultrie (Georgia) City Council are elected to staggered terms from the town at-large by a plurality of votes cast. 1 Plaintiffs, black residents of Moul-trie, allege that use of the at-large multi-member district 2 rather than single-member districts unconstitutionally dilutes their votes and the votes of all black citizens in Moultrie, who make up about 35% of the city’s population. The district court held that plaintiffs had failed in their burden of proving dilution and dismissed. We reverse and remand.

An appointment plan is not constitutionally infirm merely because it includes multimember or at-large districts. The burden is on the plaintiff to prove that such an electoral scheme unconstitutionally dilutes the votes of minority group members. See White v. Regester, 412 U.S. 755, 765-66, 93 S.Ct. 2332, 2339, 37 L.Ed.2d 314, 324 (1973); Zimmer v. McKeithen, 485 F.2d 1297, 1304-05 (CA5, 1973) (en banc), aff’d on other grounds sub nom. East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976). In a number of cases this court has explained the content of the plaintiff’s burden of proving dilution of a minority’s votes. We have indicated that four specific areas of inquiry are particularly important: (1) equality of access of minority group members to the political process; (2) whether past discrimination has the present effect of discouraging minority members’ participation in the electoral process; 3 (3) whether the governmental policy underlying the use of multimember districts is tenuous; and (4) the responsiveness of the government body in question to the needs of the minority community. See, e. g., Corder v. Kirksey, 585 F.2d 708, 712 n.8 (CA5,1978); Nevett v. Sides, 571 F.2d 209, 217 (CA5, 1978), petition for cert. filed, 47 U.S.L.W. 3247 (Sept. 22, 1978) (No. 78-492); Kirksey v. Board of Supervisors, 554 F.2d 139, 143 (CA5) (en banc), cert. denied, 434 U.S. 968, 98 S.Ct. 512, 54 L.Ed.2d 454 (1977); Zimmer, 485 F.2d at 1305.

The purpose of inquiry into these areas is to determine whether an electoral system that is unobjectionable in the abstract, nevertheless, on the specific facts presented, has the effect of diluting the impact of a minority’s voting power, and whether it is intended to have such an effect. See Nevett v. Sides, 571 F.2d at 221-25. 4 These four areas of inquiry are not exclusive, and a plaintiff need not prove that all four inquiries produce results tending to show unconstitutional discrimination. Id. at 224; Zimmer, 485 F.2d at 1305. 5 The district court correctly identified the relevant primary areas of inquiry, but, for the reasons that we now discuss, its decision cannot stand.

I. Adequacy of findings of fact

F.R.C.P. 52(a) requires the district court to make findings of fact and conclusions of law in deciding all cases tried without a jury, and these must be sufficiently detailed that the court of appeals can ascertain the factual and legal basis for the district court’s ultimate conclusion. See, e. g., Hydrospace Challenger, Inc. v. Tracor/MAS, Inc., 520 F.2d 1030, 1034 (CA5, 1975). Because the resolution of a voting dilution claim requires close analysis of unusually complex factual patterns, see Corder v. Kirksey, 585 F.2d at 712-13, and because the decision of such a case has tho potential for serious interference with static functions, see Hendrix v. Joseph, 559 F.2d 1265, 1271 (CA5, 1977), we have strictly adhered to the rule 52(a) requirements in voting dilution cases and have required district courts to explain with particularity their reasoning and the subsidiary factual conclusions underlying their reasoning. “[Cjonclusory findings as to each of the Zimmer criteria are no more helpful than an overall conclusory finding of dilution. The factual predicates for such conclusions must be clearly stated by the trial court.” David v. Garrison, 553 F.2d 923, 929 (CA5, 1977).5 6 Perhaps in no other area of the law is as much specificity in reasoning and fact finding required, as shown by our frequent remands of voting dilution cases to district courts. See, e. g., Corder v. Kirksey, supra; Blacks United for Lasting Leadership, Inc. v. City of Shreveport, 571 F.2d 248 (CA5, 1978); Hendrix v. Joseph, supra; David v. Garrison, supra; Nevett v. Sides, 533 F.2d 1361 (CA5, 1976). As a general rule, if the district court reaches a conclusion on one of the Zimmer inquiries without discussing substantial relevant contrary evidence, the requirements of rule 52 have not been met and a remand may be called for if the court’s conclusions on the other Zimmer inquiries are not sufficient to support a judgment.

II. District court’s evaluation of the evidence

The district court held that the political process in Moultrie is equally open to participation by blacks, that past discrimination does not preclude present effective participation by blacks in Moultrie’s electoral system, that the Moultrie city government is “not unresponsive” to the needs of the black community, and that the policy underlying Moultrie’s choice of at-large elections was not one of racial discrimination. We examine each of these areas to see whether the court made full nonconclusory findings of fact and conclusions of law and used correct legal standards in evaluating the evidence.

A. Denial of access to the electoral process

A key issue in a voting dilution case is whether the minority group of which the plaintiff is a member is denied equal access to the various phases of the political process, including nomination, campaigning, voter registration, and voting. If lack of minority input into the electoral process can be demonstrated then the plaintiff has gone far toward proving that the at-large electoral system has the effect of diluting minority votes. A denial of equal access may take any of several forms, ranging from such direct governmentally-sanctioned exclusions as the poll tax or the white primary, to the existence of a private slating organization that uses racist tactics and does not seek minority votes (as in White v. Regester, supra), to less concrete but no less effective barriers to participation such as cultural and language differences between the majority and minority, see White v. Regester, 412 U.S. at 768, 93 S.Ct. at 2340, 37 L.Ed.2d at 325, or a disproportion between the levels of education, income, employment, and living conditions of the majority and minority. See Kirksey, 554 F.2d at 144 & 145.

The district court found no current legal barriers (that is, barriers imposed by law) to full black participation in the Moultrie electoral process. The white primary was abolished in Georgia over 30 years ago, and blacks are currently allowed to register and vote in all elections. Elections in Moultrie are non-partisan, and a person who wishes to run for office in Moultrie need only file an intention to run and pay a filing fee in order to be placed on the ballot. The district court found affirmative evidence of blacks’ equal access to the Moultrie political process in the recent election of a black candidate to the City Council 7 and in the fact that other, but unsuccessful, black candidates apparently had received some white votes. The court concluded that the plaintiffs had failed to prove that the Moultrie political process was not equally open to participation by blacks. 8

The district court’s conclusion of no denial of access cannot stand, first because the court did not make findings concerning evidence tending to show that official discrimination that deters political participation by blacks has not yet ended in Moultrie. Plaintiffs introduced evidence that the all-white Lions Club supervises Moutrie City Council elections. They put in other evidence that after substantial numbers of blacks had begun to register and vote the City moved a polling place from a location convenient to residents of black neighborhoods to a less convenient location farther away. Under the rule 52(a) standard, as strictly applied in voting dilution cases, this seemingly relevant contrary evidence was required to be discussed and considered.

Moreover, the inquiry into equality of access should not come to a halt simply because the evidence shows that there are no longer any barriers imposed by law to minority access. Non-legal barriers may also preclude full black participation in the political process, 9 and plaintiffs intro duced substantial evidence, not discussed by the district court, tending to show the existence of such non-legal barriers: evidence that black candidates had encountered difficulties in campaigning in white neighborhoods and evidence that housing conditions, employment rates, income, and educational levels are considerably less advantageous for Moultrie blacks than for Moultrie whites. As we held in Kirksey, evidence of socioeconomic inequities gives rise to a presumption that the disadvantaged minority group does not enjoy access to the political process on an equal basis with the majority. “Inequality of access is an inference which flows from the existence of economic and educational inequalities.” Kirksey, 554 F.2d at 145. This substantial evidence tending to show inequality of access must be considered and the Kirksey presumption applied to it.

B. Effect of past discrimination on present participation in political process

We have recognized that past policies of racial discrimination may have present impact upon the participation of the minority in the political process. The mere removal of past official discrimination does not render the present effects of that discrimination irrelevant in determining whether an electoral scheme dilutes the votes of the minority. See Kirksey, 554 F.2d at 145-46; Zimmer, 485 F.2d at 1306.

In Moultrie 44.6% of the whites and 26.9% of the blacks are registered to vote. The number of black members on the City Council historically has been grossly disproportionate to the percentage of blacks in Moultrie’s population, with only two blacks elected in the history of the town. According to the record only one black sits on the six-member Council (a mayor and five Council members), although about 85% of the citizens of Moultrie are black. Plaintiffs introduced much evidence of recent and pervasive discrimination, and the district court took judicial notice of the past official discrimination. The record contains evidence that literacy tests were used in Moultrie until the mid-sixties, that voting and registration were carried out on a segregated basis until the early sixties, that Moultrie schools were fully desegregated only in 1970, and that public recreational facilities were segregated by law until 1968 or 1969.

Since until the past decade theré had been de jure discrimination in many facets of Moultrie government, including voting for public office, the inference is strong that the disproportions in voter registration and elected officials are at least in part a result of the past pervasive discrimination in Moultrie. The district court found that there was no evidence supporting such an inference and concluded instead that it is equally likely that the reason for the disproportion in voter registration is that the “leaders in the black community” have failed “to ignite the patriotic fervor of their brothers.” This misallocated the burden of proof on the issue of present effects of past discrimination. Once plaintiffs have demonstrated a history of pervasive discrimination and a present disproportion in voting registration and election of minority representatives, they have carried their burden of proving that the past discrimination has present effects. See, e. g., Hendrix v. Joseph, 559 F.2d at 1270; Kirksey, 554 F.2d at 144 & 146; Bradas v. Rapides Parish Police Jury, 508 F.2d 1109, 1112 (CA5, 1975); Zimmer, 485 F.2d at 1306. The defendants must then come forward with rebutting evidence proving that current dispropor tions are not an effect of the past. Kirk-sey, 554 F.2d at 144-46. Without a factual basis in the record a court may not simply attribute present disproportions to lack of interest or apathy on the part of blacks. Id. at 145. Defendants concede on appeal that the district court misplaced the burden of proof but contend that they introduced sufficient evidence to carry their burden of rebuttal. This is a matter to be addressed to the district court on remand.

C. Responsiveness to minority needs

The district court in a voting dilution case must consider whether the plaintiffs have proved that elected officials are unresponsive to minority needs. If so, then the plaintiffs have made a strong case that the challenged electoral scheme has had the effect of depriving minority members of equal representation, see Wallace v. House, 515 F.2d 619, 622-24 (CA5, 1975), vacated on other grounds, 425 U.S. 947, 96 S.Ct. 1721, 48 L.Ed.2d 191 (1976), and such a showing of unresponsiveness “is strongly corroborative of an intentional exploitation of the electorate’s bias.” Nevett v. Sides, 571 F.2d at 223.

The district court held that plaintiffs had not proved Council members’ unresponsiveness to the needs of the black community. The evidence introduced that pertained to the Council’s responsiveness fell into two categories: (1) the provision of governmental services to the minority community and (2) the distribution of municipal jobs and appointments to various city boards. 10

With respect to most community services provided by the City (e. g., sewage, police protection, fire protection, garbage collection), the district court found that no evidence had been introduced showing unresponsiveness. 11 The district court cited specific instances of the Council’s acting favorably on requests for action by black citizens. The court found that there are more unpaved streets in black neighborhoods than white but discounted this as evidence of current unresponsiveness because more funds are now spent on street improvements in black neighborhoods than in white neighborhoods. The court held plaintiffs’ evidence that city schools were desegregated only in 1970, that jury lists in county courts contain disproportionately low numbers of blacks, that housing patterns in Moultrie remain segregated, and that blacks are underrepresented in the county government, was irrelevant because the City Council has no responsibility for these matters and no power to remedy them.

The district court also found no evidence of unresponsiveness in the City’s employment practices, though only between 24% and 30% of the City’s employees are black (as compared to a 35% black general population) and a greater disparity exists in supervisory level City jobs. The court found this evidence did not show unresponsiveness because the City had engaged in an affirmative action hiring plan and because the City’s employment advertisements contain the phrase “equal opportunity employer.” The court concluded that the racial disparity in the makeup of the City’s work force was not attributable to a lack of responsiveness on the part of the City but to “the apparent lack of interest on the part of blacks in applying for city jobs.” The court also found affirmative evidence of Council responsiveness in the fact that some blacks had been appointed to city boards and declined to infer any lack of responsiveness from the Council’s decision to allow vacancies on the Council to remain unfilled until the next election rather than appointing blacks to serve out the unexpired terms.

Finally, relying on the testimony of present Council members and past solicitation of black votes by white Council members, the district court found that the Council is more responsive to black needs and concerns under the present at-large system than it would be if single-member districts were used, because a single-member district system “would eliminate any political incentive for the other [white] members of the Council to give active consideration to black concerns because they would be representing only a white constituency.”

The district court’s conclusion that the plaintiffs failed to prove unresponsiveness must be set aside for several reasons. First, as explained in Kirksey, once plaintiffs have demonstrated recent pervasive official unresponsiveness to minority needs, 12 the burden shifts to the defendants to demonstrate that the unresponsiveness of the past is no longer indicative of present governmental unresponsiveness. Kirksey, 554 F.2d at 144-46. 13

Second, several items of evidence not discussed by the district court must be considered pursuant to Rule 52(a). Plaintiffs introduced uncontradicted testimony that recreational facilities in the black areas of town are inferior to those in white areas. 14 They introduced evidence that public housing projects were constructed in the 1950’s that were segregated by law when built and that remain largely segregated today. Though there is no evidence that de jure segregation remains today in Moultrie public housing, if a high degree of de facto segregation continues today and is traceable to past intentional discrimination by the Council, 15 it would be evidence of current unresponsiveness. As we have noted before, the Supreme Court has used this approach in other areas of desegregation. See Kirksey, 554 F.2d at 145, n. 12, citing Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 26, 91 S.Ct. 1267, 1281, 28 L.Ed.2d 554, 572 (1971) (burden on school district to justify plan containing racially imbalanced schools, where school district has history of past intentional segregation). Appellants go further and argue if any racial inequality exists in the town that the Council has failed to remedy, a finding of unresponsiveness is compelled. This is unsound. Where the racial discrimination in question was not caused by the acts of the local governmental body involved, the body is not necessarily unresponsive to minority needs merely because, with its limited resources, it has not attempted to remedy every existing inequality.

There is other evidence, some of it temporally remote, that arguably demonstrates Council unresponsiveness and must be considered: e. g., segregated voter lists, the City’s delay in complying with the requirements of § 5 of the Voting Rights Act of 1965, operation of racially segregated jails, and racially selective enforcement of laws by City police. 16

The district court should also further address itself to whether the Council’s expenditures for paving in black neighborhoods show unresponsiveness. The court did not consider the disproportionately large number of unpaved streets in black areas as indicative of unresponsiveness because at present more is spent on street improvements, in the black community than in the white. That current expenditures are greater in black neighborhoods does not necessarily show responsiveness. 17

Finally, the conclusion that plaintiffs had failed to prove unresponsiveness cannot stand because the court relied in part on a finding that Council members would be less responsive to black concerns if elected from single-member districts than they are under the existing at-large system. Comparisons of various electoral configurations in order to ascertain which alternative most fairly assures responsiveness to black concerns and best ensures black access to the political system is appropriate when choosing the appropriate remedy for an unconstitutional electoral system. See e. g., U.S. v. Board of Supervisors, 571 F.2d 951, 956 (CA5, 1978). However, when considering whether an existing system is unconstitutional such comparisons are not called for — the crucial question is whether government officials at present are, or are not, unresponsive to black needs.

One argument advanced by plaintiffs on appeal is patently without merit and need not be considered by the district court on remand. Plaintiffs argue that any action taken by Council members out of concern for their “mere political self-interest, j. e., soliciting votes” cannot be considered responsive to the needs of the black community. It is characteristic of representative democracy that elected officials are sensitive to the concerns of the voters at least in part because they want to receive votes when they run for reelection.

Plaintiffs also argue that the district court erred in not considering the disparity in the city’s hiring to be prima facie evidence of unresponsiveness that must be rebutted by the city. Cases cited in support of this proposition involve the burden of proof in Title VII employment cases. Since a mere statistical disparity generally cannot establish a constitutional violation in constitutional cases, see Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976), we are not convinced that the wholesale importation of Title VII principles into voting rights cases based on the Fourteenth and Fifteenth Amendments is appropriate. It is particularly difficult to see how a bare statistical disparity shows a lack of responsiveness where there is a finding that the governmental body is attempting to remedy the disparity by use of an affirmative action plan. Since the case is to be remanded, however, it would be appropriate for the district court to spell out the precise content of the city’s “affirmative action” plan. It is not clear from the opinion that this plan consists of anything more than advertising the city’s current willingness to hire without regard to race. If this is the extent of the plan and the current statistical disproportion is attributable to past intentional discrimination by the city, an affirmative action plan calculated to do away with the disparity may be required in order to rebut an inference of unresponsiveness.

D. State policy

The district court found that there was no state policy favoring either single-member or multimember districting and that many Georgia municipalities use multimember districting. The court noted that the at-large scheme had been instituted in Moultrie in 1859 when no blacks voted so that no inference could be made of racially discriminatory motivation in instituting the scheme. The district court thus properly placed no weight on this factor. The purpose of the Zimmer state policy inquiry is to decide whether an inference can be drawn that the use of multimember districts is rooted in racially discriminatory motives; the reason for inclusion of “tenuous state policy” in the Zimmer listing is that a state policy in favor of at-large districts that is shown to be “tenuous” is evidentiary of invidious intent. See Nevett v. Sides, 571 F.2d at 224; Zimmer, 485 F.2d at 1307 (state policy said “tenuous” where long tradition against at-large districts reversed shortly after large members of blacks began to vote). Where the circumstances will not support an inference of invidious motivation either in the institution or continuation of the at-large system, the absence of a state-wide policy favoring at-large district-ing is not in itself evidence of dilution.

III. Other issues

The district court’s conclusion that the aggregate of the Zimmer criteria does not show dilution of the black minority’s voting power in Moultrie City Council elections obviously must be reconsidered, since we have set aside the findings on three of the four criteria. Extended discussion of other errors asserted by the plaintiffs is not necessary. Brief discussion of one claim of error may be useful to the district court on remand, however. The plaintiffs contend that the district court erroneously held that the election of a single black official forecloses any possible dilution claims, and arguably the last sentence of the district court’s opinion can be so read. Just as the election of disproportionately few minority members does not necessarily mean that the minority vote is diluted, see Zimmer, 485 F.2d at 1305, the election of a single black official does not mean that the black vote is necessarily not diluted. See Kirksey, 554 F.2d at 149 n.21.

REVERSED and REMANDED.

1 . Moultrie had a majority vote requirement and run-off elections until 1977 when a three-judge district court enjoined their further use.

2 . We need not for purposes of this opinion draw any distinctions between “at-large” and “multimember” districts. See generally Corder v. Kirksey, 585 F.2d 708, 713 n.11 (CA5, 1978).

3 . The first area of inquiry, equal access to the election process, entails investigation into whether blacks now have an opportunity to fully participate in all phases of the electoral process — nomination, campaigning, and voting. The second area of inquiry, the effects of past discrimination, entails determining whether, although past barriers to participation such as the poll tax or white primary no longer exist, the residual effect of past discrimination is that blacks in fact continue to participate proportionately less than whites (by registering to vote in low numbers, for example). See Kirksey v. Board of Education, 554 F.2d 139, 145 n.13 (CA5) (en banc), cert. denied, 434 U.S. 968, 98 S.Ct. 512, 54 L.Ed.2d 454 (1977).

4 . Nevett v. Sides settled in this circuit that invidious discriminatory intent is required in voting dilution cases based on the Fourteenth and Fifteenth amendments. This issue is pending before the Supreme Court. See Bolden v. City of Mobile, 571 F.2d 238 (CA5, 1978), prob. jur. noted, 436 U.S. 902, 98 S.Ct. 2229, 56 L.Ed.2d 399 (1978), scheduled for reargument next term, —U.S.—, 99 S.Ct. 2048, 60 L.Ed.2d 658 (1979). On remand the district court should address itself to the question of intent.

5 . This court has also identified a number of “enhancing” factors that, if present, strengthen the plaintiff’s case. These include the use of a majority vote requirement, an anti-single shot voting provision, lack of a geographic subdis-trict requirement, and the existence of large districts. See Nevett, 571 F.2d at 223; Zim-mer, 485 F.2d at 1305.

6 . See Corder v. Kirksey, 585 F.2d at 713: “Given the intensely factual nature of voting dilution cases, we, as an appellate court, can but speculate whether the law was properly applied if we lack sufficiently explicative findings.”

7 . This candidate, Frank Wilson, was the first black elected to the City Council in the history of the city. Since this appeal was filed, Wilson was defeated for reelection by another black man. No other seats on the City Council have been won by blacks.

8 . The district court also held that blacks and whites in Moultrie do not “vote strictly along color lines,” and gave examples of black candidates who had received some white votes. The district court viewed this as evidence that blacks do have some success in campaigning for office. This is perhaps evidence that blacks have some access to the political process, although it alone could not support a finding of equal access.

Evidence of racial polarization in voting has been held to be a prerequisite to a voting dilution challenge. See Nevett, 571 F.2d at 223 & n.16, citing United Jewish Organization v. Carey, 430 U.S. 144, 166 n.24, 97 S.Ct. 996, 1010, 51 L.Ed.2d 229, 246 (1977). If race plays no part in voters’ choices, there is no injury to blacks as a group caused by the use of multi-member districts. See Comment, Constitutional Challenges to Gerrymanders, 45 U.Chi.L. Rev. 845, 856 (1978). It does not appear that the district court intended to find that there is such an absence of racial bloc voting in Moul-trie that a finding of dilution is foreclosed, however. Such a finding on this record would be clearly erroneous. See Nevett, 571 F.2d at 223 n.18. No black candidate has ever received even a plurality of white votes and Wilson, the first black elected to the Council appears to have received as little as 5% of white votes.

9 . Since our opinion in Nevett, the Zimmer criteria are required to do double duty — they must show that the effect of the multimember system is to dilute minority voting power and they must show discriminatory intent in the institution or continuation of this electoral system. Some facts unearthed in the course of the Zimmer inquiries will more readily support an inference of intent than others. For example, proof of a current governmentally-sanctioned barrier to minority access to the political process (official discouragement of black voter registration, for example) would be strong evi dence, standing alone, of invidious discriminatory intent. On the other hand, proof of current socioeconomic barriers to black participation would not, standing alone, support an inference of invidious intent, see Blacks United for Lasting Leadership, Inc. v. City of Shreveport, 571 F.2d 248, 253 & n.5 (CA5, 1978), although such evidence of non-government sanctioned denial of access could support an inference of invidious intent if coupled with evidence of unresponsiveness of elected officials to minority needs. See Nevett, 571 F.2d at 223 & n.5. The dual goals of the sensitive Zimmer factual inquiry make it imperative that the district court carefully and precisely state its findings of fact and the legal conclusions it draws from those facts.

10 . Plaintiffs also introduced evidence that individual Council members and former Council members belong to racially segregated clubs and churches. Such evidence of personal actions is not relevant in determining whether an official act was undertaken with invidious discriminatory intent. See Palmer v. Thompson, 403 U.S. 217, 91 S.Ct. 1940, 29 L.Ed.2d 438 (1971).

11 . Appellants argue that the district court’s finding of responsiveness in police protection was clearly erroneous because one witness testified that the city jail was operated on a segregated basis until 1975 and there was some testimony of discriminatory enforcement of the law by the police department during the mid-sixties. Since we are remanding the case there is no need to decide whether the district court’s decision on this factual matter was clearly erroneous.

12 . A present white member of the Council testified that prior to his election in 1972 the Council had neglected the needs of the black community.

13 . The district court is, of course, entitled to consider whether past examples of unresponsiveness are so far distant in time “to permit an inference that what was true in the past is no longer true.” Kirksey, 554 F.2d at 144.

14 . The court did discuss one incident in which a request for improvements in the condition of a swimming pool in a black community had been complied with by the Council. The district court properly found this to be affirmative evidence of responsiveness, but it must be considered along with countervailing evidence of nonresponsiveness with regard to recreational facilities.

15 . It is not clear from the record before us what degree of control the City Council now has, or had in the fifties and since, over public housing.

16 . It is possible that the segregated private housing patterns in Moultrie are also attributable to past intentional discrimination by the City. The district court found that they were not, but this finding was not adequately supported by subsidiary findings of fact. On remand the district court should address this question again.

17 . For example, if 95% of the need for street improvements is for streets in the black community, allocating only 60% of street improvement funds for work in black areas would be evidence of unresponsiveness, not responsiveness. '