Summary
holding that property owners alleged injury by claiming that newly constructed housing projects would increase traffic, noise, and crime
Summary of this case from Maya v. Centex Corp..Opinion
No. 97-11083
March 16, 1999
Michael M. Daniel, Laura Beth Beshara, Dallas, TX, for Plaintiffs-Appellees and Intervenor Plaintiff-Appellee.
Linda Frances Thome, U.S. Department of Justice, Appellate Section, Civil Rights Division, David Kevin Flynn, U.S. Department of Justice, Washington, DC, for department of Housing Urban Development.
Robert E. Goodfriend, Michael P. Lynn, Thomas M. Melsheimer, Eric Wolf Pinker, Lynn, Stodghill, Melsheimer, Tillotson, Dallas, TX, for Plaintiffs-Appellants.
Joseph G. Werner, Melissa Ann Miles, Haynes Boone, Dallas, TX, for The Housing Authority of the City of Dallas.
Appeal from the United States District Court for the Northern District of Texas
The Dallas Housing Authority (DHA), the United States Department of Housing and Urban Development (HUD), and the City of Dallas were found liable several years ago for unconstitutional racial discrimination and segregation within Dallas's public housing programs. The primary issue on this appeal is the constitutionality of the provision of the district court's most recent remedial order that directs newly constructed units of public housing to be located in "predominantly white" Dallas neighborhoods.
Specifically, this is an appeal from a final judgment in two actions that were consolidated for trial. In the first action, two homeowners and their homeowners' associations ("Homeowners") sought declaratory and injunctive relief against DHA's construction of two new public housing projects adjacent to their neighborhoods. The Homeowners challenged the remedial order's provisions for new public housing construction and race-conscious site selection alleging that these were not narrowly tailored to remedy the vestiges of past discrimination and segregation. In the second action, the original class plaintiffs, tenants in the public housing programs, sought declaratory relief that the remedial order provisions are constitutional. The district court entered judgment against the Homeowners in the first action and for the class plaintiffs in the second action. The Homeowners appealed. We essentially vacate and remand for further consideration by the district court.
The two homeowners' associations are: (1) Highlands of McKamy IV and V Community Improvement Association and (2) Preston Highlands Homeowners' Association, Inc.
I. BACKGROUND
Part of the convoluted history of this case is concisely recounted in Walker v. HUD, 912 F.2d 819, 821-25 (5th Cir. 1990) [hereinafter Walker IV]. We will not repeat that history here, but some important procedural and substantive gaps in this court's prior opinion, which addressed different issues, should be filled in.
This case began in 1985 and initially resulted in a consent decree, which was approved by the district court in 1987. See Walker v. HUD, 734 F. Supp. 1231, 1247-82 (N.D. Tex. 1989) [hereinafter Walker I] (reprinting the district court's 1987 consent decree and its "Findings of Fact Conclusions of Law Approving the Proposed Consent Decree"). The consent decree addressed the plaintiff class's challenge under the Equal Protection Clause of the Fourteenth Amendment and 42 U.S.C. § 1983 to the purposeful racial discrimination and segregation within DHA's public housing programs. The defendants were DHA and HUD. The City of Dallas was joined as both a defendant to the lawsuit and a party to the consent decree in 1989. See Walker v. HUD, 734 F. Supp. 1289 (N.D. Tex. 1989) [hereinafter Walker III]. The history of public housing in Dallas is a sordid tale of overt and covert racial discrimination and segregation. See generally Walker III, 734 F. Supp. at 1293-1312 (recounting in detail the history of public housing in Dallas). Virtually all non-elderly public housing units were constructed in minority areas of Dallas. No new public housing units were built between 1955 and 1989 at least in part for fear that they might be located in white areas. Tenant selection and assignment procedures for public housing units were crafted and administered to maintain racially segregated projects. DHA's Section 8 housing programs were operated to discourage blacks from moving into white areas of metropolitan Dallas. See id. Blacks were purposefully segregated for decades into either Section 8 housing in minority areas of Dallas or predominantly black housing projects in minority areas of Dallas.
The consent decree defined the plaintiff class as "all black persons presently or who during the pendency of this Decree become either (a) residents of a DHA owned or managed project, or (b) participants in the DHA Section 8 Existing Housing Program." See Walker I, 734 F. Supp. at 1263.
The plaintiff class will be referred to throughout this opinion as the "Walker plaintiffs." Deborah Walker is one of the named plaintiffs representing the class.
"Public housing units" refer to housing units owned and operated by DHA. In contrast, under DHA's Section 8 certificate and voucher programs, participants rent housing units from private owners, and their rents are subsidized by DHA.
In 1994, of DHA's approximately 6,400 public housing units, 6,100 were in minority areas and 300 were in predominantly white areas. An additional 75 units are currently under construction or just completed in a predominantly white area (Frankford Marsh project), and the units at issue in this lawsuit would add another 80 units to predominantly white areas of Dallas.
The 1987 consent decree required the demolition of approximately 2,600 units of public housing in DHA's West Dallas project, a public housing development located in a predominantly black area of the city and referred to by this court as "one of Dallas's worst slums." Walker IV, 912 F.2d at 821. These units were to be replaced on a one-for-one basis with additional public housing units and Section 8 certificates and vouchers. See id. at 822. The decree also required that one hundred newly constructed replacement units be built in a predominantly white area of Dallas, that a nondiscriminatory tenant selection and assignment plan be implemented, and that a Section 8 mobility plan be established to assist black families joining the Section 8 program in finding housing in white areas of Dallas.
The West Dallas project was completed in 1955 and contained 3,500 units. In many ways, it is at the heart of this litigation. It was constructed to solve the "Negro housing problem." See Walker III, 734 F. Supp. at 1295. It is the second largest public housing project in the United States. See id. at 1296 n. 21. As of 1983, one-third of its units were so dilapidated as to be uninhabitable. See id. at 1307. In 1986, rejection rates for available units in the West Dallas project ranged from 58-60%. See id. at 1308. At that time, there were 1,583 vacant units of which 1,300 had been boarded up for more than ten years because of their deteriorated state. See id. The 1987 consent decree was designed to alter this concentration of public housing by requiring the demolition of all but approximately 900 of the West Dallas units, the reconfiguration and rehabilitation of the remaining units, and the creation of new public housing opportunities in predominantly white areas of the city and its suburbs.
The Section 8 mobility program was designed to educate landlords about the Section 8 program and to assist Section 8 participants in locating Section 8 housing in non-minority areas of Dallas. Specifically, the program was to include landlord recruitment, exceptions to HUD's "fair market value" rent caps on Section 8 vouchers and certificates, Title VIII enforcement actions, and child care and transportation services. The program was also to provide neighborhood specific information on crime rates, job training and employment opportunities, day care, medical facilities, neighborhood shopping, transportation, social services, objective indicators of schools quality such as TAAS test results, and "environmental hazards or other conditions inimical to family life."
DHA repeatedly violated the 1987 consent decree. First, it resisted the construction of the 100 units of new public housing in a predominantly white area. See Walker I, 734 F. Supp. at 1243-45. Site selection for and construction of the 100 units was eventually completed, but only by court order. See id. Second, DHA violated the tenant selection and assignment and mobility provisions of the decree. See id. at 1235-42. DHA failed to establish and fund the required Section 8 mobility program, failed to timely obtain fair market exception rents, delayed implementing a nondiscriminatory tenant selection and assignment program, failed to include in Section 8 housing information a full list of all Section 8 units available in non-minority areas, and failed to use all of the Section 8 certificates and vouchers allocated by HUD to DHA. See id.
The value of a section 8 voucher or certificate is capped by HUD at the "fair market rent." This cap, however, can be increased by special application to HUD.
In March 1992, the district court vacated the 1987 consent decree on the grounds that its terms were not implemented and that the vestiges of purposeful segregation persisted. Subsequently, the district court granted the Walker plaintiffs' uncontested motion for summary judgment on the issue of liability. In September 1994, the district court held a trial on the issue of a remedy. The district court entered its remedial order affecting DHA in February 1995 and its remedial order affecting HUD in April 1996.
The remedial order affecting DHA requires DHA (1) to demolish at least 2,630 units of its West Dallas project, (2) to develop 2,807 replacement units for the demolished West Dallas units through both new construction and Section 8 vouchers and certificates, (3) to develop, either through construction or acquisition, an additional 3,205 new units of public housing in predominantly white areas of metropolitan Dallas in which the poverty rate does not exceed 13%, and (4) to develop all new public housing units in predominantly white areas until there are as many units in predominantly white areas as there are in minority areas. A "predominantly white area" is defined as less than 37% Hispanic, black, or other minority. The required 3,205 new units may be satisfied by the use of Section 8 certificates or vouchers, but only after court approval. The construction costs for 674 of the 2,807 new replacement units have been previously allocated to DHA by HUD, although only 75 of these units are currently under construction or completed.
These 2,807 units have already been funded by HUD. The 2,807 units are made up of 774 new public housing units and 2,033 Section 8 certificates and vouchers. Specifically, the replacement units consist of (i) the 100 units constructed pursuant to court order from the 1987 consent decree, (ii) 1,335 Section 8 certificates and vouchers previously funded by HUD, (iii) 339 new public housing units allocated to DHA in 1990 and 1991 as part of the proposed 2,000 unit West Dallas project, (iv) 335 new public housing units allocated under HOPE VI in 1994 as replacement units for West Dallas, and (v) 698 additional Section 8 vouchers promised by HUD. Therefore, there are 674 new units of public housing currently funded but not under construction. Eighty of these units are specifically at issue in this case.
The source of this final detail of the district court's remedial order is unclear. The written order itself is confusing. Paragraph A.3 requires that all of the 3,205 new units be built in predominantly white areas.
In slight contrast, the district court reiterates throughout his oral and written opinions that all of the 674 new units of public housing ( i.e., the ones currently funded by HUD but not under construction) "and any other allocated in the future, must be developed in predominantly white areas until there are approximately as many non-elderly public housing units in those areas as in minority areas." 10/6/97 Written Opinion, at 22. The district court goes on to say, "The race conscious site selection remedy is limited. It applies only until there is a comparable number of public housing units in white and minority areas." Id. at 22-23; see also 8/25/97 Oral Opinion, at 22 ("Again, the 674 units must be placed in white non-minority areas until there are as many public housing units in white areas as in minority areas.").
To date, no plan for the use of Section 8 certificates or vouchers has been submitted to the district court. However, HUD states in its brief to this court that it "expects to meet this obligation by providing funding for Section 8 certificates and vouchers for all of the 3,205 units, at a rate of 320 units per year for 10 years. Thus, it is likely that the only public housing units developed under the Remedial Orders will be the 774 units already allocated by HUD and designated as replacement units under paragraph A.1 of the HUD Remedial Order and paragraph A.2 of the DHA Remedial Order."
The district court authorized that 200 of the replacement units may be used in the reconfiguration and revitalization of the West Dallas project.
The 75 units either under construction or completed are at DHA's Frankford Marsh site.
The Homeowners filed this suit against DHA and HUD to enjoin the construction of two new 40-unit public housing projects on sites adjacent to their neighborhoods. The Homeowners allege that the remedy of new construction is not narrowly tailored because it requires that the new units be constructed in predominantly white areas. The Homeowners do not contest either the remedial order's poverty site-selection criterion or HUD's site-selection standards set forth in 24 C.F.R. § 941.202.
One site is at the intersection of McCallum and Meandering Way, and the other site is at the intersection of Hillcrest and Highway 190.
In response to the Homeowners' action, the Walker plaintiffs sought declaratory relief that the remedial order was constitutional. The Homeowners' request for an injunction and the Walker plaintiffs' declaratory judgment action were tried together in October 1996. The district court denied the Homeowners' injunctive relief and granted the Walker plaintiffs declaratory relief. The district court gave an oral opinion on August 25, 1997, entered final judgment on September 18, 1997, and issued its written opinion on October 6, 1997.
These opinions will hereinafter be cited respectively as: 8/25/97 Oral Opinion, 9/18/97 Final Judgment, and 10/6/97 Written Opinion.
II. STANDING
As an initial matter, DHA and HUD challenge the Homeowners' standing to bring their suit. The burden of establishing standing rests with the party seeking to invoke federal jurisdiction ( i.e., the Homeowners). See United States v. Hays, 115 S.Ct. 2431, 2435 (1995); Lujan v. Defenders of Wildlife, 112 S.Ct. 2130, 2136 (1992). In a case that has proceeded to final judgment, the factual allegations supporting standing (if controverted) must be supported adequately by the evidence adduced at trial. See Hays, 115 S.Ct. at 2435; Lujan, 112 S.Ct. at 2137; Gladstone Realtors v. Village of Bellwood, 99 S.Ct. 1601, 1616 n. 31 (1979); see also Charles Alan Wright et al., 13A Federal Practice Procedure § 3531.15, at 105 (1984).
"If this were an appeal from a judgment granting the defendants' motion to dismiss for lack of standing, we would be required to accept as true all of the material allegations of the complaint. However, because we are reviewing a final judgment based upon a fully developed record, we must evaluate standing from all materials of record." Pollard v. Cockrell, 578 F.2d 1002, 1006 (5th Cir. 1978) (internal citations and quotations omitted).
The irreducible constitutional minimum of standing is composed of three elements:
First, the plaintiff must have suffered an "injury in fact" — an invasion of a legally protected interest which is (a) concrete and particularized, and (b) actual or imminent, not conjectural or hypothetical. Second, there must be a causal connection between the injury and the conduct complained of. . . . Third, it must be likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.
Hays, 115 S.Ct. at 2435 (quoting Lujan, 112 S.Ct. at 2136) (internal citation and quotations omitted). In applying these constitutional standing requirements, the Court has adopted a prudential principle that bars the adjudication of "generalized grievances" against allegedly illegal government conduct. See id. In the equal protection context, this prudential principle means that standing exists only for those persons who are personally denied equal treatment by the challenged discriminatory conduct. See id. (citing Allen v. Wright, 104 S.Ct. 3315, 3326 (1984)); see also Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 102 S.Ct. 752, 768 n. 26 (1982) (rejecting the proposition that every citizen has "standing to challenge every affirmative-action program on the basis of a personal right to a government that does not deny equal protection of the laws").
A "generalized grievance" is a harm "shared in substantially equal measure by all or a large class of citizens." Warth v. Seldin, 95 S.Ct. 2197, 2205 (1975). "The prudential principle barring adjudication of `generalized grievances' is closely related to the constitutional requirement of personal `injury in fact,' and the policies underlying both are similar." Apache Bend Apartments, Ltd. v. United States, 987 F.2d 1174, 1176 (5th Cir. 1993) (en banc).
DHA and HUD principally contend that the Homeowners lack standing because they allege an injury that is conjectural and a generalized grievance. The essence of the Homeowners' complaint is twofold: (1) they have been purposefully discriminated against because of their race ( i.e., they were intentionally singled out because of their race to accommodate two new public housing projects adjacent to their neighborhoods), and (2) this decision has inflicted or threatens to inflict specific injury including decreased property values, increased crime and population density, environmental problems, and diminished aesthetic values of the neighborhood because DHA will fail to operate and maintain the two projects properly. The remedial order's explicit racial classification alone is sufficient to confer standing on these particular homeowners. In Allen v. Wright, 104 S.Ct. 3315 (1984), the Supreme Court wrote regarding the "stigmatizing injury" caused by racial discrimination that "[t]here can be no doubt that this sort of non-economic injury is one of the most serious consequences of discriminatory government action and is sufficient in some circumstances to support standing." Id. at 3326. The Court continued, "Our cases make clear, however, that such injury accords a basis for standing only to `those persons who are personally denied equal treatment' by the challenged discriminatory conduct. Id.; see also City of Richmond v. J.A. Croson, 109 S.Ct. 706, 721 (1989) ("To whatever racial group these citizens belong, their `personal rights' to be treated with equal dignity and respect are implicated by a rigid rule erecting race as the sole criterion in an aspect of public decisionmaking.").
Under the remedial order, DHA selected the homeowners' neighborhood because they are white and they live in an area of Dallas that is at least 63% white. The remedial order also requires that new units not be located in areas where the poverty rate exceeds 13%. Thus, these homeowners' "whiteness" is one of two controlling elements which identified the specific sites adjacent to their neighborhoods for new public housing construction. When a homeowner's neighborhood adjoins a proposed public housing project whose site was determined by a race-conscious standard, he has standing to sue because of the explicit racial classification. Cf. Hays, 115 S.Ct. at 2436 ("Where a plaintiff resides in a racially gerrymandered district, however, the plaintiff has been denied equal treatment because of the legislature's reliance on racial criteria, and therefore has standing to challenge the legislature's action.").
DHA and HUD cite three cases to support their contention that the Homeowners' injury is a generalized grievance lacking the specificity and particularity necessary to confer standing. Each case is easily distinguishable from the case at hand. In Warth v. Seldin, 95 S.Ct. 2197, 2208-09 (1975), the plaintiffs alleged that an adjacent town's zoning ordinances effectively excluded low and moderate income persons from living in the town, but they could not demonstrate that the ordinances specifically precluded them from living in the adjacent town. In Apache Bend Apartments v. United States, 515 F.2d 1174 [ 987 F.2D 1174], 1177 (5th Cir. 1993), the plaintiff-taxpayers were not seeking to litigate their own tax liability, but the tax liability of taxpayers who were not before the court. And in United States v. Hays, 115 S.Ct. 2431, 2436 (1995), the plaintiffs were denied standing to challenge a reapportionment plan because they did not live in the district that was the focus of their claim. In contrast to these three cases, the Homeowners live in neighborhoods next door to the proposed new 40-unit housing projects, and the location of these projects was selected specifically because of the homeowners' race.
In general, the racial classification of the homeowners is an injury in and of itself. See Shaw v. Reno, 113 S.Ct. 2816, 2824 (1993) ("Classifications of citizens solely on the basis of race are by their very nature odious to a free people whose institutions are founded upon the doctrine of equality. They threaten to stigmatize individuals by reason of their membership in a racial group and to incite racial hostility." (internal citations omitted)). But the Homeowners also allege that constructing two new 40-unit public housing projects adjacent to their neighborhoods will cause a decline in their property values and other problems involving crime, traffic and diminished aesthetic values. Relying on the district court's fact finding that the Homeowners have not suffered such an economic injury caused by the new public housing projects, HUD and DHA challenge whether the Homeowners have in fact suffered a decrease in property values. Despite this finding, we cannot conclude, having reviewed the record, that the Homeowners did not put forth adequate evidence at trial to confer standing upon them. The district court did not hold that the Homeowners lack standing, as he was well aware of the potential for neighborhood disruption traceable to improperly managed public housing projects. HUD and DHA cite no cases in which standing has been denied to homeowners who asserted their quality of life and property values would be diminished by a next-door public housing or other HUD project. The caselaw is to the contrary.
See Alschuler v. HUD, 686 F.2d 472, 476-77 (7th Cir. 1982); South East Lake View Neighbors v. HUD, 685 F.2d 1027, 1034-35 (7th Cir. 1982); Society Hill Civic Ass'n v. Harris, 632 F.2d 1045, 1054 (3d Cir. 1980).
In sum, the Homeowners alleged and sufficiently proved facts that were adequate to support standing to sue. The injury they assert is not too abstract or conjectural. The line of causation between the alleged unconstitutional conduct and the injury is not attenuated. And the prospect of obtaining relief from the injury as a result of a favorable ruling is not speculative.
As a subsidiary matter, DHA contends that the Homeowners lack standing because their alleged injury would not be redressed by the invalidation of the remedial order's requirement that public housing be built only in "predominantly white areas." In other words, DHA argues that it could have selected the two sites in question even without the race-conscious site selection criterion.
DHA applies the wrong legal standard. The test is whether DHA would have selected these two sites absent the race-conscious criterion, not whether it could have selected the sites. See Warth, 95 S.Ct. at 2208 (framing the standing question as whether, absent the challenged zoning ordinances that allegedly excluded low-income persons from living in the town of Penfield, there was a substantial probability that the plaintiff would have been able to obtain such housing). There is no evidence in the record that DHA would have chosen the two sites in question absent the race-conscious criterion, and the Walker plaintiffs admit in their brief to this court that "[t]he record is clear that DHA would not have chosen the sites absent the court order."
Additionally, HUD argues that the Homeowners are challenging the wrong part of the remedial order by contesting the 3,205 additional units of public housing that must be built in predominantly white areas. It is quite clear from the record, however, that the Homeowners have properly challenged the two 40-unit projects, which were selected on a race-conscious basis to be built next to their neighborhoods. Obviously, these two projects are part of a larger remedial scheme that is affected by this court's opinion herein.
III. HOMEOWNERS' EQUAL PROTECTION CLAIM
The district court held that the Homeowners failed to allege an equal protection violation. The Homeowners challenge this conclusion.
A. Lack of a Similarly Situated Group
The district court, citing Samaad v. City of Dallas, 940 F.2d 925 (5th Cir. 1991), found that the Homeowners failed to show an equal protection violation because they did not identify a similarly situated set of black persons who have been treated better. See id. at 941 n. 31 This requirement, however, applies only to equal protection claims involving facially neutral government actions, where it is necessary to establish that the government is distinguishing or classifying persons on the basis of race. See id. at 941. Explicit racial classifications, in contrast, establish unequal treatment by their very nature. See Shaw v. Reno, 113 S.Ct. 2816, 2824 (1993) ("Laws that explicitly distinguish between individuals on racial grounds fall within the core of [the Equal Protection Clause's] prohibition."). Because the Homeowners challenge an explicit racial classification within the district court's remedial order, they have properly alleged an equal protection violation.
The district court also suggested that the Homeowners failed to demonstrate an equal protection violation because "[t]he impact [of the new construction], if any, on the Homeowners will be considerably less than [the] impact of the existing DHA public housing projects on the property owners in the black neighborhoods with existing projects." The district court's reasoning is incorrect: racial classifications are not acceptable simply because they are perceived to have little impact. Any explicit racial classification, regardless of the burdens or benefits its imposes, is suspect and subject to strict scrutiny. See Adarand Constructors, Inc. v. Pena, 115 S.Ct. 2097, 2113 (1995) ("[A]ll racial classifications, imposed by whatever federal, state, or local government actor, must be analyzed by a reviewing court under strict scrutiny."); Powers v. Ohio, 111 S.Ct. 1364, 1370 (1991) ("It is axiomatic that racial classifications do not become legitimate on the assumption that all persons suffer them in equal degree.").
B. Intent to Discriminate
The district court held that the Homeowners failed to prove an equal protection violation because there is no intent to treat whites worse than similarly situated blacks. Once again, the district court is incorrect. An explicit racial classification does not require any inquiry into "intent" in order to allege an equal protection violation. See Shaw, 113 S.Ct. at 2824 ("No inquiry into legislative purpose is necessary when the racial classification appears on the face of the statute."). "Express racial classifications are immediately suspect because, `[a]bsent searching judicial inquiry . . ., there is simply no way of determining whether classifications are `benign' or `remedial' and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.'" Id. (quoting Richmond v. J.A. Croson, 109 S.Ct. 706, 721 (1989)).
The district court's skepticism of the Homeowners' right to pursue an equal protection claim was unfounded, so we proceed to address the merits of their claim.
IV. NARROW TAILORING
The primary issue on appeal is whether the remedial order's requirement that new public housing units be built or acquired in "predominantly white areas" is narrowly tailored to remedy the vestiges of past discrimination and segregation within Dallas's public housing programs. No party suggests on appeal that the racial steering, which for so many years was a part of Dallas's public housing programs, has not been effectively eradicated. Rather, the parties' dispute centers on the district court's efforts to rectify the effects of the now-past discriminatory practices of DHA, HUD, and the City of Dallas. In short, is it constitutional in this case to implement a race-conscious site selection criterion for newly built or acquired public housing?
See 8/25/97 Oral Opinion, at p. 5 ("From the very beginning the primary purpose that DHA had was to prevent blacks from moving into white areas of the city and the suburbs. That is not true of DHA today. . . .").
Any race-conscious remedial measure receives strict scrutiny under the Equal Protection Clause. See Adarand Constructors, Inc. v. Pena, 115 S.Ct. 2097, 2113 (1995); Black Fire Fighters Ass'n v. Dallas, 19 F.3d 992, 995 (5th Cir. 1994) [hereinafter BFFA]. This is true no matter which race is burdened or benefitted by the racial classification in question. See Adarand, 115 S.Ct. at 2110-11 (citing City of Richmond v. J.A. Croson Co., 109 S.Ct. 706, 722 (1989)). Strict scrutiny requires that a racial classification be (1) justified by a compelling government interest and (2) narrowly tailored to further that interest. See Adarand, 115 S.Ct. at 2113. The Homeowners do not contest that there exists a compelling government interest in this case. Therefore, our inquiry focuses on whether the remedial order is narrowly tailored.
In assessing whether a remedy is narrowly tailored, courts are to assess five factors: (1) the necessity for relief, (2) the efficacy of alternative remedies, (3) the flexibility and duration of relief, (4) the relationship of the numerical goals to the relevant market, and (5) the impact of the relief on the rights of third parties. See United States v. Paradise, 107 S. Ct. 1053, 1066 (1987); BFFA, 19 F.3d at 995. Before examining these factors, however, we must address the standard of review.
When a district court's race-conscious remedial measure is challenged as not being narrowly tailored, the party defending the remedial measure bears the burden of producing evidence that the remedial measure is constitutional. See Wygant, 106 S.Ct. at 1848-49; Aiken v. City of Memphis, 37 F.3d 1155, 1162 (6th Cir. 1994); see also Raso v. Lago, 135 F.3d 11, 20 (1st Cir. 1998) (Stahl, J., dissenting). The party challenging the remedial measure, of course, bears the ultimate burden of demonstrating that the racial classification is unconstitutional. See id.
In general, we review a district court's legal conclusions de novo and its findings of fact for clear error. However, "if the trial court bases its findings upon a mistaken impression of applicable legal principles, the reviewing court is not bound by the clearly erroneous standard." Inwood Labs., Inc. v. Ives Labs, Inc., 102 S.Ct. 2182, 2189 n. 15 (1982); see also United States v. Bentley-Smith, 2 F.3d 1368, 1373 (5th Cir. 1993).
In the case at hand, the district court placed the burden of production, as well as the ultimate burden of proof, on the Homeowners. Because he improperly placed the burden of production, we are not bound by the clearly erroneous standard in reviewing his findings of fact. Nonetheless, although more rigorous review may be in order, we will deferentially examine the district court's findings because this is a complicated case in which the district court has a decade's worth of experience with Dallas's public housing programs.
Race-conscious remedies must be narrowly tailored to eliminate the effects of past discrimination as well as bar like discrimination in the future. See Paradise, 107 S.Ct. at 1067-68; id. at 1073 (citing Louisiana v. United States, 85 S.Ct. 817, 822 (1965)). "Racial classifications are simply too pernicious to permit any but the most exact connection between justification and classification." Adarand, 115 S.Ct. at 2113 (quoting Fullilove v. Klutznick, 100 S.Ct. 2758, 2805 (1980) (Stevens, J., dissenting)). This means that a race-conscious remedy must be framed to address the exact effects and harms of the discrimination at issue. See Wygant, 106 S.Ct. at 1850.
In application, arriving at an exact fit between harm and remedy requires consideration of whether a race-neutral or less restrictive remedy could be used. See Adarand, 115 S.Ct. at 2118; Wygant, 106 S.Ct. at 1850 n. 6. This is because a race-conscious remedy should be the remedy of last resort. See Alexander v. Estepp, 95 F.3d 312, 316 (4th Cir. 1996) ("[E]xplicit racial preferences, if available at all, must be only a `last resort' option."). A "race-conscious remedy will not be deemed narrowly tailored until less sweeping alternatives — particularly race neutral ones — have been considered and tried." Williams v. Babbitt, 115 F.3d 657, 666 (9th Cir. 1997).
It is true that in Hills v. Gautreaux, the Supreme Court did not disapprove a remedial order which, to end discrimination in public housing, required housing to be built in nonblack neighborhoods of Chicago. Hills, 425 U.S. 284, 296, 96 S.Ct. 1538, 1546 (1976). But the Court did not consider the propriety of that remedy as opposed to a non-race conscious remedy like a vigorous Section 8 program. The only disputed issue in that case was whether the remedy could extend beyond the city limits of Chicago. Further, Hills predates significant changes that have occurred both in HUD's approach to public housing and in the scrutiny afforded race-conscious remedies. Hills does not mandate either the construction of new public housing or race-conscious site selection. We must review the application of the Paradise factors, which the district court correctly understood to frame the issues here.
The first two Paradise factors are the necessity for relief and efficacy of alternative remedies. These weigh against race-conscious site selection for two reasons. First, Section 8 housing vouchers have not been given a fair try to prove their potential to desegregate. Second, other criteria than a racial standard will ensure the desegregated construction or acquisition of any new public housing. See infra, text at n. 31.
The Homeowners argue that section 8 alone is capable of remedying the effects of past discrimination in Dallas's public housing programs, and they contend that section 8 has not been given a fair try. They also point out, and the district court agrees, that Section 8 is more cost-efficient than new construction and is preferred by the majority of public housing program participants. Section 8 is more flexible than fixed public housing because the participants may decide where and in what type facility to reside. As a result, virtually all available Section 8 vouchers have been snapped up in Dallas.
Unfortunately, however, numerous programs that would encourage and assist black families to use section 8 in predominantly white areas had been in effect or fully operational for only a short time before trial on the Homeowners' case. For instance, the record indicates that DHA's mobility program had been operating as originally proposed by the Walker plaintiffs and the district court since only approximately 1994. In addition, exceptions to the "fair market rent" caps on section 8 vouchers and certificates have been slow in coming. It also appears that potential section 8 landlords may now be paid "signing bonuses" for accepting section 8 tenants. These are but three examples of numerous "helping-hands" that may be employed to promote the success of section 8 as a desegregation tool. While the history and timing of their implementation by DHA and HUD in this case are not perfectly clear, it is evident from the record that they were not programs of long-standing before trial. If section 8, combined with such assistance programs, is an effective desegregation tool, then section 8 is superior to a race-conscious remedy in that it allows market forces and personal preferences rather than racial criteria to guide the homemaking decision.
See 10/30/96 Testimony of Ann Lott, Director of DHA's Section 8 Mobility Program (agreeing that most of the efforts to improve DHA's mobility program have been within the preceding two years); 8/7/96 Deposition of Lott (explaining new procedures within DHA's mobility program that were implemented in the prior year, including a marketing campaign for prospective landlords, a landlord newsletter, and private briefings for landlords on the section 8 program). But see 10/30/96 Testimony of Lori Moon, President CEO of DHA (stating that DHA's mobility efforts have been in effect since 1989).
It is unclear from the record whether HUD has granted DHA a 120% fair market exception rent across-the-board for all section 8 housing in predominantly white areas. It is also unclear whether DHA could be granted fair market rent exceptions higher than 120%. Evidence in the record shows that a 160% fair market exception rent would significantly increase the availability of section 8 housing in predominantly white areas. Prior to implementing a race-conscious remedy, it would seem advisable to ensure (1) that rent exceptions above 120% are foreclosed, as they offer the potential to break-down a significant barrier to an even more effective section 8 program, or (2) that 120% fair market exception rents have been fully implemented and tried, if that is the statutory maximum.
Another example would be the vigorous enforcement of state and federal laws prohibiting racial discrimination by private section 8 landlords.
The district court found, agreeing with the Walker plaintiffs, DHA, and HUD, that section 8 needed to be combined with new construction or acquisition in predominantly white areas in order to remedy the effects of past discrimination. Adopting the Walker plaintiffs' proposed findings of fact and conclusions of law in toto, the court concluded that Section 8 alone was an inadequate remedy for several reasons: there are not enough Section 8 units in predominantly white areas; among the available units there is a lack of three and four bedroom units; rents in predominantly white areas are too high to be covered even by Section 8's fair market exception rents; landlords do not want to participate in the Section 8 program; and Section 8 participants become frustrated in looking for housing in predominantly white areas and settle for housing in minority areas. The court also found that rental contract requirements in predominantly white areas contain provisions that are difficult for Section 8 families to meet ( e.g., high security deposits, requirement of having held a job for the past year, etc.). We neither accept nor attempt to reject these factual findings. Rather, there is one overarching factual finding by the district court — which is uncontested by all parties — that transcends the parties' objections to Section 8 as a remedial measure.
In 1987, when the district court first found DHA in violation of its original consent decree, a negligible number of black families in DHA's Section 8 programs lived in predominantly white areas. In contrast, in September 1994, approximately 1,050 Section 8 black families lived in predominantly white areas. And in the fall of 1996, there were approximately 1,335 Section 8 black families in predominantly white areas. These numbers show that in the two year period between 1994 and 1996, the number of Section 8 black families living in predominantly white areas increased by 285, or 27%. And it was during this same time period that DHA's Section 8 mobility program was getting fully underway. The program currently in place, which the district court has not criticized, essentially became fully operational around 1994. Based on the relative success of DHA in moving blacks into predominantly white areas via its Section 8 program between 1994 and 1996, the Walker plaintiffs, HUD, and DHA have produced insufficient evidence to show that the district court's race-conscious site selection criterion is necessary to remedy the effects of past discrimination. Cf. In re Birmingham Reverse Discrimination Employment Litig., 20 F.3d 1525, 1545-47 (11th Cir. 1994); see also BFFA, 19 F.3d at 995 ("The broad skip promotion remedy in the decree is difficult to justify when the knowledge to narrow it seems readily available."). Section 8, a race-neutral remedial measure, is increasingly successful at moving black families into white areas, and the record indicates that it could be even more successful with, for example, increased funding for both more vouchers and the mobility program, more mobility counselors, and higher fair market exception rents. When Section 8 has evidenced such promising results, options such as these should be explored and tested before adopting a race-conscious remedy as a last resort.
The Walker plaintiffs state that in 1987 there were 66 Section 8 black families living in predominantly white areas, which amounted to approximately 2.4% of all black families in DHA's Section 8 program.
According to testimony in the record, this amounted to approximately 21% of all black families in DHA's Section 8 program.
According to testimony in the record, this amounted to approximately 24-25% of all black families in DHA's Section 8 program.
In In re Birmingham Reverse Discrimination Employment Litigation, non-black employees of the Birmingham Fire Rescue Service (BFRS) challenged a 1981 consent decree entered into by the City of Birmingham that required, inter alia, that 50% of all promotions to the rank of BFRS lieutenant would be filled by qualified blacks ( i.e., for every two promotions to lieutenant, one must be filled by a black candidate). See id.
The 11th Circuit noted that between 1978 and 1981, BFRS significantly increased its total number of black employees from 8 to 42, which represented a shift in the black percentage of BFRS's workforce from 1.89% to 9.3%. See id. This was achieved without the use of a race-conscious remedy. See id. The court stated that "there is strong evidence in the record of this case that the City had implemented effective alternatives to race-based quotas to remedy its prior discriminatory behavior." Id. "While the district court correctly concluded that, when the decree was entered, no black had as yet become a fire lieutenant, we believe that, given the City's progress at the entry-level, alternative measures designed to increase black representation in the fire lieutenant ranks were feasible." Id. The court proceeded to list the non-race-based remedies that the plaintiffs proposed regarding promotion to BFRS lieutenant. See id. at 1546-47. It then concluded, "Considering the efficacy of the alternative remedies, the relief provided in the decree cannot be reconciled with the requirement that a government's use of race must be narrowly tailored." Id. at 1547.
See 10/30/96 Testimony of Ann Lott, Director of DHA's Section 8 Mobility Program, at III-43 (stating that DHA is able to use virtually every Section 8 voucher or certificate funded by HUD); see id. at III-65 (stating that DHA could put to use an additional 1,500 Section 8 vouchers).
See id. at III-67 (stating that the mobility program could use more counselors).
See id. at III-57 (stating that higher fair market exception rents would assist in obtaining more Section 8 housing in predominantly white areas).
Additionally, it is unnecessary to employ the race-conscious site selection criterion ordered by the court even if new construction or acquisition of public housing occurs. The district court may proceed with new construction as part of his remedial plan, but he may not do so using a race-conscious site selection criterion. Of course, other criteria may be employed to guide site selection for new construction. See United States v. Yonkers Bd. of Educ., 837 F.2d 1181, 1184, 1236-37 (2d Cir. 1987) (using a geographical site selection criterion for public housing). The district court has already mandated that all new sites be in areas where the poverty rate does not exceed 13%. The district court's concern that if it does not attach a race-conscious site selection criterion to new construction, then the new units will end up in minority areas and, as a consequence, Dallas's public housing projects will almost all remain in minority areas, is unfounded.
It is suggested in brief that this criterion essentially restricts the areas available for new construction to predominantly white areas.
Moreover, the district court referred throughout his 8/25/97 Oral Opinion and 10/6/97 Written Opinion to the danger of losing the funds already allocated to new construction because they could not be transferred to the Section 8 program. This may be true. Nonetheless, it does not justify the use of a race-conscious site selection criterion.
In deciding that Section 8 and nonracial site selection criteria should be implemented before a racial standard, we are mindful of the "respect owed a district judge's judgment that specified relief is essential to cure a violation of the Fourteenth Amendment." Paradise, 107 S.Ct. at 1073. Nonetheless, a district court's broad equitable powers remain constrained by the boundaries of narrow tailoring. The recent success of Section 8 and the availability of viable nonracial, non-discriminatory site selection criteria, combined with the factors discussed in the following paragraphs, demonstrate that the remedial order's race-conscious site selection criterion is not narrowly tailored.
See Billish v. City of Chicago, 989 F.2d 890, 893 (7th Cir. 1993) ( en banc); Wessmann v. Gittens, 160 F.3d 790, 808 (1st Cir. 1998) (" Croson, in particular, leaves no doubt that only solid evidence will justify allowing race-conscious action . . ."); id. ("Our dissenting brother's valiant effort to read into Croson a broad discretion for government entities purporting to ameliorate past discrimination strikes us as wishful thinking.").
First, DHA, HUD, and the City of Dallas are all cooperating defendants. That is to say, DHA, HUD, and the City of Dallas no longer discriminate against black families in DHA's public housing programs, and — by all accounts in the record — all three defendants are active participants in crafting and implementing remedial measures to eliminate the vestiges of past discrimination. In contrast, the Supreme Court approved a race-conscious remedy in United States v. Paradise in large part because earlier, less restrictive remedies had proven ineffective since the defendant continually resisted their implementation and stonewalled in developing acceptable procedures for the advancement of black troopers within the Alabama Department of Public Safety. See Paradise, 107 S.Ct. at 1062-63 (describing the Department's continuing failure to comply with the parties' consent decrees); id. at 1066, 1069; see also BFFA, 19 F.3d at 996 (finding a race-conscious remedial measure unnecessary, in part, because the defendant was a willing party to the settlement of the lawsuit). Where, as here, the defendants have begun making race-neutral, good faith, and effective efforts to remedy the wrongs of the past, a race-conscious remedy should only be a last resort.
Second, the district court, in his 8/25/97 Oral Opinion, made references to the necessity of a race-conscious site selection criterion because some participants in DHA's public housing programs do not want and should not be forced to use Section 8. While some may find it difficult to use Section 8, a race-conscious remedial measure is not justified by certain class members' objections to looking for housing on their own versus their being offered a unit owned and operated by DHA. A race-conscious remedy is justified, after race-neutral remedies have been considered and found wanting, if it is the only effective means by which to remedy the effects of past discrimination. It is by this standard alone that the district court must assess his remedial orders. As applied to the facts of this case, the district court's concern seems particularly irrelevant as only 474 new units in predominantly white areas are currently funded, and those units must be filled by Section 8 families who participate in DHA's Family Self-Sufficiency Program. That is, they will be filled by the "cream-of-the-crop" from DHA's waiting lists. Thus, those who in all probability need public housing units the least (because they would be successful Section 8 participants) will be directed to the new units, while those for whom the district court expresses concern will be left with Section 8. DHA's rationale for filling the new units with participants in its self-sufficiency program is commendable, but it throws askew part of the district court's reasoning regarding the remedial need for new construction in addition to Section 8.
See, e.g., 8/25/97 Oral Opinion, at p. 28 ("Even if this money [for new public housing construction] could be reprogrammed for Section 8, there would still be a need for the use of public housing in the desegregation plan. Section 8, although the preferred method, is not the only method. Indeed, there are members of this class who made it evident, abundantly clear to this Court, that they did not trust Section 8, they would not use Section 9, and they should not be forced to use Section 8."); id. at p. 33 ("I recognized then, as I do now, that although Section 8 certificates and vouchers may be the preferred method for most people, for many people there are risks in Section 8 that not all class members would want to take. I declined to force those risks upon unwilling class members.").
DHA's Family Self-Sufficiency Program is a voluntary program that requires participants to, for instance, be employed or attend school. The participants sign a five-year contract with DHA in which they agree to abide by the program's regulations. In essence, the program is designed make its participants self-sufficient by providing them with "upward mobility type assistance." 8/7/96 Deposition of Lori Moon, President CEO of DHA, at p. 110.
Third, it remains unclear why there is an absolute remedial necessity to build 474 new units of public housing using a racial classification when (1) 2,033 (or 72%) of the remedial order's 2,807 replacement units will be Section 8 and (2) DHA and HUD may submit a plan to use Section 8 for all of the 3,205 additional units of public housing to be built in predominantly white areas. In sum, out of a total of 6,012 units within the remedial order, only 474 (or 8%) must — assuming an acceptable Section 8 plan is submitted for the 3,205 additional units — be new construction in predominantly white areas. If Section 8 can effectively satisfy the district court's remedial goal regarding 5,628 units, it is baffling to assume that it cannot do so for an additional 474 units.
As noted previously, HUD states in its brief to this court that it intends to fulfill its entire obligation regarding the 3,205 additional units with Section 8 vouchers or certificates.
DHA, HUD, and the City of Dallas offer two responses to this criticism. First, they argue that 474 units is a tiny proportion of the overall number of units contemplated by the remedial order; the deference due a district court in fashioning a remedial order should protect such a small element of the overall remedial plan. This would be correct if the 474 units were not attached to a racial classification which requires that they be built in predominantly white areas. Racial classifications, even small ones, receive strict scrutiny. Second, they contend that if the 474 units already funded by HUD are not built in white areas, they will either not be built at all or will be built in minority areas which will only further the racial segregation of DHA's public housing projects. As discussed previously, the district court may entertain any number of site selection criteria regarding new public housing units, except for race. No one suggests that the 474 units should not be built, only that requiring that they be built using a race-conscious site selection criterion is not narrowly tailored.
The fourth Paradise factor, the relationship of numerical goals to the relevant markets, also cuts against the race-conscious site-selection criterion in the remedial order. The district court's remedial goal is to have half of the families in DHA's public housing programs (either public housing units or Section 8) in predominantly white areas of Dallas and half in "minority areas". The justification for this goal is that Dallas's population is approximately half white and half "minority" and, therefore, public housing should be divided accordingly. This goal is overly broad. This suit was brought by black plaintiffs on behalf of a class of black plaintiffs. There is no suggestion that the suit ever expanded to include all minorities or that any liability of the public agencies to other minorities could be found. The court's definition of a "predominantly white" neighborhood, with 63% white population, is also based on the idea that public housing may not be placed in neighborhoods with higher concentrations of Hispanics. There is no evidence in the record to support the court's arbitrary definition of a predominantly white neighborhood. The emphasis should instead be directed toward placing public housing participants in neighborhoods of their choice through a vigorous Section 8 program, non-black neighborhoods, census tracts in which no public housing currently exists, or non-poor neighborhoods.
The third Paradise factor considers the flexibility and duration of relief. On balance, that factor is neutral in this case.
According to the Walker plaintiffs' brief, at 11: "There are 113 county census tracts with 63% or greater non-Hispanic white population, 95 of which have a poverty population less than the county average. There are an additional 20 tracts with a non-Hispanic white population between 50% and 63%. Only seven of these tracts have a poverty rate less than the county's. There are an additional 16 tracts with a poverty population below the county average but the non-Hispanic white population is less than 50%."
The fifth Paradise factor is the impact of a racially-conscious site selection criterion on the rights of third parties. Among all the groups affected by Dallas public housing, only these Homeowners have maintained that they would be injured by the racially-based site selection process that occurred here. Despite the court's having purportedly found against them on this issue, the totality of the remedial order is far more ambivalent. The district court ordered stringent criteria for the design and upkeep of the projects and for tenant selection here and in another "predominantly white" neighborhood (the Frankford Marsh site, fn. 12 supra), and he called for the participation of neighboring community members, like these Homeowners, in planning the projects. The court showed considerable sensitivity to the fact that public housing has in the past been disgracefully neglected in Dallas. The resulting remedial order thus cuts both ways with respect to the Homeowners' rights. On one hand, it attempts to placate their fears of deterioration in their neighborhoods. On the other hand, it lends credibility to those fears.
Because there are promising, non-racially discriminatory ways to continue desegregating public housing in Dallas, the provision of the court's remedial order calling for the construction or acquisition of units of public housing in "predominantly white" areas is unconstitutional. Under the balance of the Paradise factors, the criterion is not narrowly tailored, and it is premature to utilize such a last-resort measure. We must vacate and remand this portion of the remedial order for further consideration.
In so doing, we emphasize several points. First, increased reliance on Section 8 demands that the public agencies implement a vigorous mobility plan that serves the relocation needs and concerns of black families, reaches out to white landlords, affords adequate fair market rent exceptions, and combats illegal private discrimination. Second, this opinion does not deal with the remedial order's nondiscriminatory tenant selection and assignment provisions, which are not challenged by the Homeowners. Third, this opinion does not preclude the construction or acquisition of additional public housing if sites are selected by means of nonracial criteria. But we also recognize that Section 8 is overwhelmingly preferred by public housing families, that it allows market forces and personal preferences to control the homemaking decision, and that it has not proven ineffective at desegregating Dallas' public housing programs when combined with a vigorous mobility program. As applied to the facts of this case, the district court erred in employing a race-conscious remedy before utilizing race-neutral alternatives.
V. CONCLUSION
For the foregoing reasons, the district court's remedial order is VACATED to the extent indicated and the case is REMANDED for further proceedings; the declaratory judgment awarded to the Walker plaintiffs is REVERSED; and this court's stay of construction at the sites adjacent to the Homeowners' subdivisions shall be ENFORCED until the district court holds additional hearings and enters a remedial order revised in accordance with the foregoing opinion.
Remedial order VACATED and REMANDED; declaratory judgment for Walker plaintiffs REVERSED; stay ENFORCED pending entry of revised remedial order.