800 F.2d 1381 | 5th Cir. | 1986
Amy HANSON, et al., Plaintiffs-Appellants,
v.
The VETERANS ADMINISTRATION, et al., Defendants-Appellees.
No. 85-2618.
United States Court of Appeals,
Fifth Circuit.
Sept. 29, 1986.
James D. Crawford, Douglas N. Candeub, Philadelphia, Pa., for plaintiffs-appellants.
Marilyn S.G. Urwitz, Robert S. Greenspan, Attys., Dept. of Justice, Washington, D.C., for defendants-appellees.
Appeals from the United States District Court for the Southern District of Texas.
Before WISDOM, DAVIS and JONES, Circuit Judges.
W. EUGENE DAVIS, Circuit Judge:
Plaintiffs appeal the dismissal of their claims against the Veterans Administration (VA) alleging racial discrimination in its appraisal practices. We affirm.I. BACKGROUND
The VA is authorized by the Veteran's Benefits Act, 38 U.S.C.A. Secs. 101-5228 (West 1979), to operate a Home Loan Guaranty Program that allows veterans to obtain home loans under favorable terms including little or no down payment. Under the program, the VA induces qualified private lenders to loan eligible veterans 100% of a home's purchase price by guaranteeing up to 60% of the loan or $27,500, whichever is less.1 38 U.S.C.A. Sec. 1810(c) (West Supp.1986).
Before guaranteeing a loan, the VA hires a fee appraiser to appraise the home being purchased and determine its "reasonable value." The reasonable value is calculated by the "market approach" method; the fee appraiser locates three homes which were sold recently and which are as similar as possible to the property being appraised. The sales price of these three "comparables" is used as a starting point to appraise the subject property; the appraiser may adjust the sales price of the comparable property either up or down depending on the perceived differences between the comparables and the home being appraised. In using this appraisal method, the fee appraiser does not take into consideration a sales price that has been agreed to on the home being appraised.
The appraised value submitted by the fee appraiser is then reviewed by a VA review appraiser who can adopt, modify or reject the estimate. 38 C.F.R. Sec. 4340(b) (1985). When he completes his review, the review appraiser issues a Certificate of Reasonable Value (CRV) which acts as a ceiling above which the VA may not guarantee any loan. 38 U.S.C.A. Sec. 1810(b)(5) (West 1979). Additionally, the veteran must certify that he paid in cash "from his own resources" any difference between the cost of his home and its reasonable value. 38 C.F.R. Sec. 4336(a)(3) (1985). The veteran, however, can request an increase in the CRV amount by submitting additional comparables.
Each appellant2 was party to a different agreement for the sale of property in which a VA guaranteed home loan was sought.3 The properties at issue in all of the agreements are located in the MacGregor subdivision, a predominantly black, middle class neighborhood, in Houston, Texas. In each instance, the VA's appraisal estimate and CRV value were below the price prospective purchasers had agreed to pay for the property. These "underappraisals," as they are referred to in the industry, caused many of the prospective purchasers to reduce their offers or look elsewhere for a home where a 100% VA loan was available.4
Appellants filed suit against the VA seeking an injunction and damages on the ground that the value of their property was discriminatorily adjusted downward by the VA fee appraisers because MacGregor was a racially mixed neighborhood. As support, appellants offered statistical evidence comparing the percentage of VA underappraisals in MacGregor with that of South Hampton, a white neighborhood allegedly similar to MacGregor. The evidence purported to show that MacGregor had a significantly higher percentage of underappraisals than South Hampton; appellants argued that the underappraisals resulted from the application of racially discriminatory appraisal practices by the VA in violation of the Fair Housing Act of 1968, 42 U.S.C.A. Secs. 3601-3612 (West 1977), the Civil Rights Act of 1866, 42 U.S.C.A. Secs. 1981 & 1982 (West 1981), the Fifth Amendment, and the Thirteenth Amendment.
After a bench trial, the district court dismissed the suit on multiple grounds including a lack of standing, prescription, failure to state a claim and failure to establish discriminatory intent or effect. The district court also dismissed the damage claims on grounds that the VA enjoyed sovereign immunity as to those claims. We find it necessary to reach the following issues: (1) Do appellants have standing to sue the VA; (2) did appellants state a claim; (3) did the district court err in denying relief to appellants on the merits.5
II. STANDING
A.
As a threshold matter, we must consider appellants' standing to bring this suit. The requirement that a litigant have standing to institute litigation in federal court has evolved from two sources: (1) Article III of the Constitution limits the jurisdiction of federal courts to actual "Cases and Controversies;" and (2) courts have imposed additional limitations on litigants predicated on prudential considerations. Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 471, 102 S.Ct. 752, 757, 70 L.Ed.2d 700 (1982). The Article III prong of the standing doctrine requires the plaintiff to "allege personal injury fairly traceable to the defendant's allegedly unlawful conduct and likely to be redressed by the requested relief." Allen v. Wright, 468 U.S. 737, 751, 104 S.Ct. 3315, 3325, 82 L.Ed.2d 556 (1984). This test requires the court to examine (1) whether the alleged injury is "distinct and palpable;" (2) the causal connection between defendant's conduct and the alleged injury; and (3) the causal connection between the alleged injury and the relief sought. Id. at 751, 753 n. 19, 104 S.Ct. at 3325, 3326 n. 19.
Prudential considerations which have been incorporated into the standing doctrine include "the general prohibition on a litigant's raising another person's legal rights, the rule barring adjudication of generalized grievances more appropriately addressed in the representative branches, and the requirement that a plaintiff's complaint fall within the zone of interests protected by the law invoked." Id. at 751, 104 S.Ct. at 3325. These prudential concerns do not, however, apply to suits brought under sections 810 and 812 of the Fair Housing Act, 42 U.S.C. Secs. 3610, 3612, respectively. Congress intended standing under these statutes to be as broad as Article III allows; therefore, courts are without authority to apply their self-imposed restrictions on standing in such suits. Gladstone, Realtors v. Village of Bellwood, 441 U.S. 91, 109, 99 S.Ct. 1601, 1612, 60 L.Ed.2d 66 (1979). See also Havens Realty Corp. v. Coleman, 455 U.S. 363, 372, 102 S.Ct. 1114, 1120, 71 L.Ed.2d 214 (1982). Because they raise a claim under these statutes, our inquiry into appellants' standing is limited to the requisites of Article III.
B.
Applying the Article III standard, we are persuaded that at least one plaintiff, Professor Otis King, has standing. King contracted to buy the house at 4304 Roseneath for $46,950. He was unable to purchase the house because the VA issued a CRV for $40,000 thereby limiting the amount of the loan they would guarantee, and he was unable to raise $6,950 from his own resources for a down payment. King later agreed to purchase the house at 4106 South MacGregor Way for $55,000. The VA appraised this home and issued a CRV for $51,000. King was able to pay a $4,000 down payment and bought the house. King alleges that racially discriminatory appraisal practices resulted in the VA's underappraisals of these homes which prevented him from acquiring the Roseneath home.
The district court found that the VA's low appraisal of the Roseneath home was based on non-discriminatory factors, such as the presence of termites, dry rot and foundation settlement, and not the racial mix of the neighborhood. The court determined that King's inability to obtain the $6,950 down payment resulted in his withdrawal from the sale and that this "inability cannot be traced to any wrongful act of Defendants." Regarding the South MacGregor house, the district court found that King failed to demonstrate a distinct injury.
We disagree with the district court's conclusion. King undoubtedly suffered a "distinct and palpable" injury, because he was unable to purchase the Roseneath home. See Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 262, 97 S.Ct. 555, 561, 50 L.Ed.2d 450 (1977). The district court apparently found that Professor King's injury--his inability to purchase the Roseneath home--was not related to the VA's conduct because King was unable to prove that the VA's appraisal practices were illegal.
This reasoning misconstrues the purpose and elements of standing. "In essence the question of standing is whether the litigant is entitled to have the court decide the merits of the dispute or of particular issues." Warth v. Seldin, 422 U.S. 490, 498, 95 S.Ct. 2197, 2205, 45 L.Ed.2d 343 (1975). It is inappropriate for the court to focus on the merits of the case when considering the issue of standing. O'Hair v. White, 675 F.2d 680, 685 (5th Cir.1982). In examining the causation element of standing, the question to be asked is whether the line of causation between King's injury and the VA's allegedly biased appraisal method is "too attenuated." Allen, 468 U.S. at 752, 104 S.Ct. at 3325.
We answer this question in the negative. King's inability to purchase the Roseneath home is "fairly traceable" to the low VA appraisal. The VA is barred from guaranteeing a loan in excess of the CRV amount and the regulations require King to raise any additional money necessary for the purchase "from his own resources." Thus, the VA underappraisal had a direct effect on King's ability to purchase the Roseneath home.
The district court also concluded that King failed to establish a causal connection between his injury and requested relief because he was not entitled to money damages by law and an injunction would not "trigger any increases in the property values at issue." Here too the district court's reasoning is based on its conclusion that the VA's appraisals were properly conducted. "The redressability element of Art. III is designed to bar disputes which will not be resolved by judicial action." Community Nutrition Institute v. Block, 698 F.2d 1239, 1249 (D.C.Cir.1983), rev'd on other grounds, 467 U.S. 340, 353 n. 4, 104 S.Ct. 2450, 2458 n. 4, 81 L.Ed.2d 270 (1984). The question to be asked is whether "the prospect of obtaining relief from the injury as a result of a favorable ruling is too speculative?" Allen, 468 U.S. at 752, 104 S.Ct. at 3325.
For purposes of standing we conclude that the relief sought--raising the VA appraisals by enjoining discriminatory appraisal practices--might permit Professor King to purchase the Roseneath home. Also, if appraisals are raised in the MacGregor subdivision generally, the value of homes owned by Professor King and others in the subdivision may increase.
We, therefore, conclude that King did allege a "personal injury fairly traceable to the [VA's] allegedly unlawful conduct and likely to be redressed by the requested relief." Id. at 751, 104 S.Ct. at 3325. Because Professor King has standing to bring this action, we need not consider the standing of the remaining appellants. Watt v. Energy Action Educational Foundation, 454 U.S. 151, 160, 102 S.Ct. 205, 212, 70 L.Ed.2d 309 (1981); Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. at 264 n. 9, 97 S.Ct. at 562.
III. FAILURE TO STATE A CLAIM
The district court held that appellants failed to state a claim for relief under section 804(a) of the Fair Housing Act which makes it unlawful "(a) to refuse to sell or rent after the making of a bona fide offer or to refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex or national origin." 42 U.S.C. Sec. 3604(a) (1977). The district court held that because the plaintiffs did not complain of discrimination in the sale or rental of housing their complaint failed to state a claim under the Act.
The authorities do not support the district court's conclusion. Courts have consistently given an expansive interpretation to the Fair Housing Act; to state a claim under the Act, it is enough to show that race was a consideration and played some role in a real estate transaction. Moore v. Townsend, 525 F.2d 482 (7th Cir.1975).
In United States v. Mitchell, 580 F.2d 789 (5th Cir.1978), the district court found that the defendant, which owned an apartment complex, steered black tenants to a particular section of the complex and that this effectively denied the black tenants access to equal housing opportunities. We affirmed the conclusion of the district court that these acts by the defendant made unavailable or denied "a dwelling to any person because of race." We held that "steering evidences an intent to influence the choice of the renter on an impermissible racial basis. The government need only establish that race was a consideration and played some role in the real estate transaction." 580 F.2d at 791 (citations omitted).
We conclude that section 804(a) does address the claim asserted by appellants. Discriminatory appraisal may effectively prevent blacks from purchasing or selling a home for its fair market value. This interferes with the exercise of rights granted by the Fair Housing Act.
IV. THE MERITS
A finding of intentional racial discrimination is necessary for recovery against a defendant under the Civil Rights Act and the Fifth Amendment.6 Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 567, n. 16, 104 S.Ct. 2576, 2590 n. 16, 81 L.Ed.2d 483 (1984) (Sec. 1981); Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976) (Fifth Amendment); Save our Cemetaries, Inc. v. Archdiocese of New Orleans, Inc., 568 F.2d 1074, 1078 (5th Cir.1978), cert. denied, 439 U.S. 836, 99 S.Ct. 120, 58 L.Ed.2d 133 (1978) (Sec. 1982). However, a violation of section 804 of the Fair Housing Act may be established not only by proof of discriminatory intent, but also by a showing of a significant discriminatory effect. Woods-Drake v. Lundy, 667 F.2d 1198, 1202 (5th Cir.1982); United States v. Mitchell, 580 F.2d 789, 791-92 (5th Cir.1978). See also Arthur v. City of Toledo, Ohio, 782 F.2d 565, 574 (6th Cir.1986) (and cases cited therein).
The district court found that appellants failed to prove discriminatory intent or effect. The question for decision narrows to whether the subsidiary findings of fact essential to this ultimate determination are clearly erroneous. Irby v. Sullivan, 737 F.2d 1418, 1424 (5th Cir.1984).7
We take note of the Supreme Court's recent instruction that:
If the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact it would have weighed the evidence differently. Where there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous.
Anderson v. City of Bessemer City, North Carolina, 470 U.S. 564, 573-74, 105 S.Ct. 1504, 1512, 84 L.Ed.2d 518 (1985). We conclude, after careful examination of the record, that the district court's findings are not clearly erroneous.
A. Discriminatory Intent
Appellants first assert that the district court did not recognize the significance of their evidence in reaching its conclusion that the VA appraisals and CRV's at issue conformed to the established "market approach" method of appraisal without discriminatory intent.
Appellants introduced proof that a widely used appraisal text, The Appraisal of Real Estate, instructed appraisers until 1977 that the value of the property being appraised should be adjusted downward if the ethnic composition of the neighborhood to which it belonged was not homogeneous. The "principal of conformity" categorized different ethnic groups according to their detrimental effect upon property values after their "infiltration" into the neighborhood.
Although this principle was no longer taught after 1977, appellants produced evidence that the VA had never issued guidelines instructing its appraisers not to use this racially biased practice and argued that the VA's appraisers, who had learned the "principal of conformity," continued to apply it. They contend that this is borne out by evidence that several VA appraisal reports concerning MacGregor property referred to "economic depreciation," "changes in the neighborhood" and "lack of pride of ownership," all of which their experts testified indicated racial considerations.
In response to this evidence, five Houston appraisers examined the appraisals in evidence and testified that these phrases did not have racial connotations. They also testified that the appraisal values appeared reasonable and that they were not aware of any Houston appraiser who used racial consideration in their appraisals, emphasizing that such consideration would be unethical. Additionally, the VA appraisal forms require the appraiser to certify that he did not take race into account in the appraisal.
We are persuaded that the district court was entitled to accept the testimony of the VA's experts over that of appellants' experts. The district court's decision "to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, ... if not internally inconsistent, can virtually never be clear error." Anderson, 470 U.S. at 575, 105 S.Ct. at 1513.
Appellants also contend that the district court failed to properly appreciate the significance of the evidence that the VA appraisals of MacGregor property were riddled with errors, such as understating the size of a home or overstating its age. They argue that these numerous errors--all of which tended to justify a lower appraisal--considered collectively, create an inference of racial bias. The district court did not err in characterizing these as "independent instances of human error" which did not prove discriminatory intent. This finding is supported by testimony of several Houston appraisers that correction of some of the errors would not necessarily entitle the property owner to a higher appraisal.
B. Discriminatory Effect
Appellants argue that the discriminatory effect of the VA appraisals is established by their statistical evidence. The statistical evidence they presented is based on a comparison of facts surrounding sales of homes in MacGregor and South Hampton subdivisions in which VA financing was sought. Appellants obtained the addresses of the homes being sold, the sales prices agreed to by the parties and the VA appraised value of the homes. Based on this information, appellants' expert, Dr. Barton Smith, made computations which, he testified, revealed that the VA appraisals discriminated against prospective buyers and sellers of MacGregor homes.
Dividing the samples into two time periods, 1970-1977 and 1978-1982, Smith found that 80% and 86%, respectively, of the MacGregor homes were underappraised. The 80% figure was based on twenty-four samples and the 86% figure was based on thirteen; Smith found a 2% chance that this many underappraisals would occur on a random basis. Smith concluded that the four sample sales in South Hampton used for 1970-1977 was too small to be reliable, but based on seven sales in the 1978-1982 period he found that 29% of the VA appraisals in South Hampton were underappraisals.
Smith also calculated that for the periods 1970-1975 and 1976-1980, respectively, the rate of appreciation of the home values in MacGregor was 6.8% and 15.7% while the average appreciation rate for Houston was 11.2% and 13.9%. He testified that this was unusual because MacGregor was no longer experiencing the effects of the racial transition which ended in approximately 1970. He concluded that the homes in MacGregor should have appreciated at a greater rate than the market generally because MacGregor was recovering from deflationary forces such as panic selling that occurred during racial transition.
Additionally, Smith performed a regression analysis on sixteen properties located in MacGregor using data compiled by the Society of Real Estate Appraisers (SREA). This analysis predicted the value of the homes assuming the market took racial factors into account (race included) and their values assuming the market excluded these factors from consideration (race excluded). He then compared these predicted values to the VA appraised values and concluded that on the average, the VA appraised value of these homes was below both predicted values. Because the VA appraisal values were below even the race included values, Smith concluded that VA appraisers were double counting the racial factors.
Based on these statistical studies, Smith determined that home prices in MacGregor were artificially being held below their true market price and expressed confidence that this resulted from the VA appraisers' improper consideration of racial factors in the neighborhood when appraising the homes. Smith further testified that the VA underappraisals could effectively place a "lid" on the entire MacGregor market. If the appraisal caused the home to be sold at a price below its true market value, its later use as a comparable for another piece of property would result in a lower appraisal, and a lower sales price, of that home. Since VA financing accounts for 45% of the institutional lending in MacGregor, the sale of one house below its true market value could have a substantial impact on the entire neighborhood.
Despite this evidence, the record supports the district court's decision to reject "the various hypotheses and conclusions offered by the plaintiffs' expert witnesses." The VA's expert, Dr. Cooke, examined the statistics offered by Smith and determined that they were inconclusive. Cooke testified that Smith failed to adequately control the data. No comparative data from other parts of Houston or comparative data on the appraisal behavior of other loans, such as FHA loans, was used as a control for the MacGregor data. In addition, Cooke found that the only comparative data available for use by Smith, South Hampton sales with VA appraisals, was derived from too few samples--a total of eleven sales from 1970-1982--to be reliable. Smith himself admitted that the lack of comparative data concerning other types of loans created a problem with his "lid" theory.
Cooke asserted that Smith had failed to consider important non-racial variables, such as crime rate and school quality, in his regression analysis although an earlier study Smith had done, using the same SREA data, found these variables to be contributing factors in the decrease of home prices. Cooke testified that the regression analysis or "hedonic model" used by Smith presumes that the effect of the omitted non-racial variables on home value is purely random. The district court was entitled to conclude that these omitted variables are important to home value and that their impact would not be purely random. If the variables of crime rate and school quality are not random in their impact on home value, Smith's analysis was skewed and this gave the district court adequate grounds to reject Smith's conclusion based on this study.
Cooke also testified that Smith's race excluded value has minimal importance in explaining how the market reacts because Smith improperly computed this value by merely removing the coefficient assigned to race from his race included value rather than recomputing the entire regression without the race variable. Cooke testified that if done properly, the race excluded value would probably not be much higher than the race included value, indicating that race is not as important a factor in the lower property values as Smith's figures suggest. Cooke also determined that some of the variables used in the regression analysis had been given incorrect values. Correcting these mistakes, Cooke reproduced the regression and determined that the new race included predicted value was statistically insignificant from the VA appraisal value, suggesting that there was no double counting of racial factors by the VA appraisers.
Cooke also recalculated the appreciation rate of homes in MacGregor and produced lower rates than Smith. Cooke testified that his lower rates weakened Smith's "lid" theory by indicating that the ceiling on prices may not be as binding as Smith's numbers suggested. He also disputed Smith's assertion that the VA underappraisals acted as a "lid" on MacGregor property values, pointing out that the data indicated that only eleven of the MacGregor underappraisals had been used as comparables from 1966-1982.
Based on all of this evidence, the district court was entitled to conclude that appellants failed to show that the VA appraisals resulted in a racially-based negative impact on home value in the MacGregor area.V.
Having concluded that the district judge was not clearly erroneous in finding appellants' evidence insufficient to establish discriminatory intent or effect, we need not review the other reasons offered by the district court for dismissing appellants' claims on the merits. The judgment of the district court is
AFFIRMED.
For the time period relevant to this suit, the maximum amount the VA was authorized to guarantee was $17,500. See Act of October 18, 1978, Pub.L. No. 95-476, Sec. 105(a), 92 Stat. 1499 (1978) (substituting $25,000 for $17,500 as the ceiling for VA loan guarantees). The $25,000 ceiling was later increased to $27,500. See 38 U.S.C.A. Sec. 1810(c) (West Supp. 1986)
A motion for class certification filed by the original plaintiffs, Amy and William Hanson and the West MacGregor Protective Association, was denied by the district court with the provision that it could be urged again at trial. After the district court denied class certification, several current or former neighborhood homeowners were joined as additional plaintiffs
Appellant West MacGregor Protective Association was not involved in a sale of MacGregor property but joined as plaintiff to protect the rights of its members who were involved in such sales
Amy and William Hanson contracted to sell their house at 4406 Roseneath for $59,000, but the property was appraised by the VA at $55,000. Judge Bonnie Fitch agreed to sell her home at 5019 Ventura for $55,000, but the VA appraised her property at $49,350 "as is" and $52,750 if she repaired its cracked slab. Judge Robert Anderson agreed to buy the house at 4114 Fernwood for $54,000, but the property was appraised at $41,500. Donald and Annie Hill agreed to purchase a home at 5324 Calhoun for $75,000, but the property was appraised at $72,750. Professor Otis King contracted to buy the property at 4304 Roseneath for $46,950, but the home was appraised at $40,000. Unable to fulfill this contract, he later agreed to buy the home at 4106 South MacGregor Way for $55,000. The VA issued a CRV on this property for $51,000
The VA conceded that it was subject to suit on appellant's claim for injunctive and declaratory relief. Because we must reach the merits to review the denial of this relief and because we ultimately find the district court did not err in denying injunctive and declaratory relief to appellants, it is unnecessary for us to consider the district court's conclusion that appellants' damage action against the VA is barred by sovereign immunity. Also, our disposition on the merits makes it unnecessary to consider the prescription issue
Appellants do not appeal the denial of their claim under the Thirteenth Amendment
Appellants contend that in their Fair Housing Act claim the district court (1) only considered whether the VA appraisals resulted in a significant discriminatory impact on appellant West MacGregor Protective Association (WMPA) but not the impact on the individual plaintiffs, and (2) that in considering the discriminatory effect on WMPA the court misconstrued the nature of the impact alleged; therefore, they argue that we are not restrained by the "clearly erroneous" standard in our review of this claim
Appellants' first argument is predicated on the district court's finding on the issue of discriminatory impact which is given in the same paragraph in which it discusses its findings regarding WMPA. The district court stated that: "Similarly, they have failed to show, by a preponderance of the evidence, a racially-based negative impact or effect caused by the Defendants' actions." A reading of the entire paragraph convinces us that the district judge did not intend to limit the finding at issue to WMPA. Our conclusion is bolstered by his specific reference to "they," which implies all appellants as opposed to WMPA alone.
Appellants' second argument is based on the court's finding that "WMPA alleges that it has been injured by Defendants because supposedly equivalent houses were sold for more money in other areas of Houston than in [MacGregor]." Appellants fault this statement on the grounds that the discriminatory impact "is not that the homes in the MacGregor neighborhood have not, in simple and absolute terms, sold for the same prices as equivalent homes in white neighborhoods." Appellants contend that the impact they alleged is that their "property values were and continue to be artificially restrained from rising to the rate at which they would otherwise rise if left to the operation of normal market forces, in contrast to the property values in comparable white neighborhoods." Our review of the record, in particular, exchanges between the district judge and the appellants' expert witness, Dr. Barton Smith, persuades us that the district judge understood the alleged disparate impact to be as articulated by appellants.
