This litigation challenges New York City's decision to close Sydenham Hospital, one of its 17 municipal hospitals, on the ground that the City's proposed action would constitute racial discrimination in the use of federal funds in violation of Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d et seq. Like most American cities, New York has struggled mightily to provide adequate municipal services with limited financial resources; its difficulties have become particularly severe since its budget crisis that began in the mid-1970's. Closing Sydenham is one of the many painful steps that the City has undertaken or proposed in an effort to maintain financial stability. The discrimination claim in this case arises from the fact that Sydenham, located in central Harlem, serves a population that is 98% minority (Black and Hispanic). Three related cases have been brought to prevent the closing of the Hospital, or at least to ameliorate the effects that the closing would have on the minority population it serves. This consolidated appeal is from the denial of a preliminary injunction. We affirm because we agree with the District Court that there is no likelihood that plaintiffs will prevail on the merits.

In April, 1979, Mayor Koch appointed a Health Policy Task Force to examine ways of reducing costly excess hospital capacity while maintaining access to high quality health services. The Task Force report, issued June 20, 1979, recommended a series of steps that the City's Health and Hospital Corporation (HHC) estimated would save $30 million in fiscal year 1981. With respect to the 17 hospitals of the municipal hospital system, the report proposed that some hospitals be replaced, that some hospitals reduce the number of beds, and that two hospitals, Sydenham and Metropolitan, both located in Harlem, be closed. The HHC approved the report on June 28, 1979. On August 12, 1979, the first of the three cases in this litigation, Bryan v. Koch, was filed on behalf of a class of low income Black and Hispanic residents of New York City who use the municipal hospital system. Defendants are the City and State of New York, the HHC, the State Health Department, and various city and state officials, including Mayor Koch. The U. S. Department of Health and Human Services (HHS) (formerly Department of Health, Education and Welfare) was joined as a defendant, though not charged with any violation of law. On September 12, 1979, the second suit, District Council 37 v. Koch, was filed by District Council 37, American Federation of State, County and Municipal Employees Union, AFL-CIO, and five of its members on behalf of its Black and Hispanic members who use the municipal hospital system. Defendants are the City, the HHC, and the city officials named in Bryan. Both suits, which have been consolidated, allege that the City's proposed plan for the municipal hospital system violates various federal civil rights statutes, primarily Title VI. In addition, the Bryan complaint alleges a pendent state law claim not involved on this appeal.

On January 25, 1980, the City defendants gave notice to the State Health Commissioner of intention to close Sydenham in 90 days. Plaintiffs moved for a preliminary injunction to prevent the closing pending the lawsuit or at least until there was adequate assurance of alternate in-patient and emergency care for the minority population served by Sydenham. The District Court for the Southern District of New York (Abraham D. Sofaer, Judge) held hearings on this motion for 13 days during March and April.

On April 30, 1980, the third suit, Boyd v. Harris, was filed on behalf of a class of low income minority residents of New York City who use the municipal hospital system. Defendants are the City and Mayor Koch, the HHC and its president, and the Secretary of HHS. The Boyd suit repeated the Title VI claims of the Bryan suit and in addition alleged that the Secretary of HHS was violating her Title VI duties by failing to investigate an administrative complaint concerning the proposed hospital closing and failing to take enforcement action against the City defendants. The Boyd complaint also alleged that the City defendants were obstructing an HHS Title VI investigation. The Boyd plaintiffs also sought a preliminary injunction to prevent the closing of Sydenham, agreeing to accept the Bryan hearing record for their injunction motion.

Until May 13, 1980, the Government had taken no position on any aspect of the litigation. On that day, Judge Sofaer, diligently endeavoring to ready the preliminary injunction motion in all three lawsuits for a prompt decision in light of the Mayor's announced decision to close Sydenham on May 16, inquired of counsel for HHS as to the federal government's position with respect to the preliminary injunction. The following day counsel for the Government advised the Court by letter that it favored the granting of a preliminary injunction, because it agreed with the Title VI allegations advanced in the Bryan suit and with the contention in the Boyd suit that the closing of Sydenham should be deferred at least until completion of HHS's Title VI investigation.

On May 15, 1980, Judge Sofaer denied the injunction in all three suits in a thorough opinion, which he amended on May 23. 492 F. Supp. 212. Judge Sofaer reached essentially four conclusions. First, he concluded that Title VI, like the Equal Protection Clause of the Fourteenth Amendment, condemns only conduct motivated by "a racially discriminatory purpose." Washington v. Davis, 426 U.S. 229, 240, 96 S. Ct. 2040, 2047, 48 L. Ed. 2d 597 (1976). Second, he found that the evidence wholly failed to show that the City's decision to close Sydenham was racially motivated. Third, he found that even if a Title VI violation could be established by the "effects" test evidence of a disproportionate racial impact or effect unjustified by any legitimate governmental purpose, see Lau v. Nichols, 414 U.S. 563, 94 S. Ct. 786, 39 L. Ed. 2d 1 (1974) the evidence adequately established the City's justification for closing Sydenham. Finally, he concluded that the availability of adequate alternative treatment for "most, if not all" of the persons served by Sydenham eliminated the prospect of irreparable harm required for issuance of a preliminary injunction.

The plaintiffs in all three suits appeal from the denial of the injunction. The Bryan and District Council 37 plaintiffs argue that the proper standard for assessing a claim of discrimination under Title VI is the effects test of Lau v. Nichols, supra, and not the intent or purpose test of Washington v. Davis, supra. Applying the effects test, they contend that the burden of closing Sydenham falls disproportionately on Blacks and Hispanics and that the closing will cause significant adverse effects because alternative health care arrangements are not assured. They further contend that the City's justification for closing Sydenham fails to satisfy Title VI, especially for lack of adequate consideration of alternative ways of achieving what plaintiffs allege would be comparable or even greater savings. The Boyd plaintiffs contend, in addition, that the closing of Sydenham should be stopped at least pending the completion of HHS's administrative investigation.

Though HHS is a defendant in the litigation, it does not appear in this Court either as appellant or appellee. However, the United States filed a brief as amicus curiae, and supplemented its views after oral argument.1II.

The standard for review of the grant or denial of a preliminary injunction as "a general rule" is whether there has been "a clear showing that the District Judge abused his discretion." Jack Kahn Music Co. v. Baldwin Piano & Organ Co., 604 F.2d 755, 758 (2d Cir. 1979). We are inclined to examine the injunction denial somewhat more rigorously on this appeal for two reasons. First, the District Court conducted nearly a plenary hearing, resolving with virtual finality the merits of plaintiffs' claim that the closing of Sydenham violates Title VI. Second, the closing is a significant event, unlikely to be altered once the action has been taken.

On the merits of the Title VI claim, we recognize the possibility of applying the intent standard of Washington v. Davis, supra. Indeed, three days after this appeal was heard, another panel of this Court declared that discriminatory intent is required whether challenged governmental action is evaluated "under the Equal Protection Clause or under the Civil Rights Act of 1964, Title VI" because the "standard of discrimination in Title VI is the same standard the Supreme Court establishes for discrimination under the Fifth and Fourteenth Amendments." Lora v. Board of Education, 623 F.2d 248, 250 (2d Cir. 1980). If we were to apply an intent standard, the District Court's denial of an injunction would clearly have to be affirmed, since the findings that the City acted without discriminatory intent are fully supported by the evidence.

However, it is at least arguable that Lora does not determine the issues before us. The arguments would be as follows. First, the Supreme Court's decision in Lau v. Nichols, supra, applying an effects standard to Title VI, has not been overruled, see Lora v. Board of Education, supra, 623 F.2d at 251 (Oakes, J., concurring), and has been viewed as controlling by this Circuit, Board of Education v. Califano, 584 F.2d 576, 589 (2d Cir. 1978), aff'd, 444 U.S. 130, 100 S. Ct. 363, 62 L. Ed. 2d 275 (1979), and other courts. Guadalupe Organization, Inc. v. Tempe Elementary School District No. 3, 587 F.2d 1022, 1029 n.6 (9th Cir. 1978); Serna v. Portales Municipal Schools, 499 F.2d 1147, 1154 (10th Cir. 1974); see Wade v. Mississippi Cooperative Extension Service, 528 F.2d 508 (5th Cir. 1976); Shannon v. HUD, 436 F.2d 809, 816-17 (3d Cir. 1970). Second, Lora, like its predecessor in this Circuit, Parent Ass'n of Andrew Jackson High School v. Ambach, 598 F.2d 705 (2d Cir. 1979), involves a court-ordered school desegregation remedy for which Title IV, 42 U.S.C. § 2000c-6 (1976), provides the relevant standard, confining a court's remedial authority to instances of de jure segregation. See 598 F.2d at 715-17. Third, even if Title VI requires proof of discriminatory intent before federal funds may be terminated, it does not necessarily follow that such a finding is required in a private litigant's suit seeking an injunction to prevent future discrimination. Finally, HHS regulations impose an effects test, explicitly prohibiting a recipient of federal funds from using "criteria or methods of administration which have the effect of subjecting individuals to discrimination," 45 C.F.R. § 80.3(b)(2), and HHC has contractually bound itself to observe these regulations. See Fullilove v. Klutznik, --- U.S. ----, 100 S. Ct. 2758, 65 L. Ed. 2d 902 (1980) (plurality opinion).

We express no view on the merits of any of these contentions, nor do we find it necessary to decide whether singly or together they provide a sufficient basis for distinguishing Lora. Even under the effects standard, we conclude that the judgment of the District Court should be affirmed. Applying that standard, we agree with the plaintiffs that they have sufficiently shown a disproportionate racial impact to require justification by the defendants. Disparity appears from comparing the 98% minority proportion of the Sydenham patients with the 66% minority proportion of the patients served by the City's municipal hospital system. Whether the impact of this disparity is sufficiently adverse to create a prima facie Title VI violation is a closer question. The District Court was satisfied with the City's estimates for the care of Sydenham patients in nearby municipal and voluntary hospitals. Nevertheless, the District Court acknowledged that at least a small number of patients, those admitted to the emergency room because of gunshot or knife wounds or drug overdoses, would suffer adverse consequences if the nearest emergency room treatment available were at even slightly more distant locations. Moreover, we share the plaintiffs' concern that the City's estimates for alternative care of Sydenham's patients rests on projections about the availability of bed space without sufficient assurance that the voluntary hospitals on which the City relies will admit all of Sydenham's patients in financial need. Whether these hospitals will admit Sydenham's Medicaid patients remains an issue in some dispute, and even if Medicaid patients will be admitted, it is also unclear what will happen with those eligible for Medicaid who are unable to establish their eligibility at the time hospital admission is needed.2

We therefore consider it appropriate to complete an assessment of plaintiffs' Title VI claim by examining the justification advanced by the City for closing Sydenham. As the District Court found, that justification is both the reduction of expenditures and the increase in efficiency within the municipal hospital system. Though the plaintiffs dispute the amount of savings the City claims will be achieved, they acknowledge that there is sufficient evidence to support the District Court's finding that closing Sydenham will reduce expenditures to some extent. However, saving money, while obviously a legitimate objective of any governmental plan to close a public facility, cannot be a sufficient justification in a case like this where public officials have made a choice to close one of 17 municipal hospitals. In such circumstances it is the choice of this particular hospital that must be justified.3

To provide a basis for making this choice the Task Force report initially assessed each of the municipal hospitals against four sets of criteria: (a) hospital size, scope of patient services, and extent of usage; (b) patient access to comparable alternative facilities; (c) quality of plant and operations; and (d) present and predicted fiscal performance. Among the several hospitals considered deficient in this assessment, recommendations for closure were made for those hospitals with disproportionately high operating deficits and obsolete plants, located within 30 minutes of other municipal hospitals with comparable or broader services. These criteria are reasonably related to the efficient operation of the City's municipal hospital system, and the evidence abundantly justifies the selection of Sydenham based on these criteria. Sydenham is the smallest of the City's 13 acute care municipal hospitals. In 1979 the average daily in-patient caseload was only 93, and the average daily emergency room caseload was only 70. The operating deficit was high, and the difference between per patient cost and Medicaid reimbursement was the highest of all the municipal hospitals. Built in 1925, its plant is obsolete and badly in need of costly renovation. Moreover, Sydenham is located within 30 minutes of another municipal hospital, Harlem, which had a 75% occupancy rate in 1979, as well as five voluntary hospitals.

If any of the municipal hospitals are to be closed, plaintiffs do not dispute that Sydenham is an appropriate choice for closing. Their claim is that the closing of a federally funded facility resulting in a disproportionate racial impact violates Title VI unless the defendants establish the unavailability of alternative measures that would save equivalent money with less disproportionate impact. Proceeding from this premise, plaintiffs further contend that instead of closing any municipal hospitals, the City could save as much money or more and avoid a disparate racial impact by such alternatives as hospital mergers, regionalization of services, increasing Sydenham's services to reduce its deficit, or increasing Medicaid reimbursement.

Neither Title VI nor the HHS regulations explicitly require a federal fund recipient to consider alternatives to a proposed placement or closing of a public facility. It is unlikely that challenges to such governmental actions were even contemplated when Title VI was enacted in 1964. The focus at that time was on federally aided facilities that denied access to minorities or admitted them only to segregated portions of a facility. See, e. g., H.R.Rep.No.914, 88th Cong. (1964), additional views of Hon. William M. McCulloch et al., reprinted in (1964) U.S.Code Cong. & Admin.News, pp. 2487, 2511. The argument for consideration of alternatives in placing or closing facilities stems from two sources. HHS has expressed its view as the administrator of federal medical care funds that its regulations should be interpreted to require consideration of alternatives. On March 5, 1980, the Undersecretary of what was then HEW wrote to Mayor Koch, in connection with the proposed closing of municipal hospitals, that if closings were to create a disparate racial impact, the City would have an opportunity to establish that they are "necessary to achieve legitimate objectives unrelated to race or national origin and that these objectives cannot be achieved by other measures having a less disproportionate adverse effect." In addition to this administrative interpretation, an "alternatives" test has also been applied by courts assessing discriminatory actions challenged under Title VII, e. g., Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S. Ct. 2362, 2375, 45 L. Ed. 2d 280 (1975) ("If an employer does then meet the burden of proving that its tests are 'job related,' it remains open to the complaining party to show that other tests or selection devices, without a similarly undesirable racial effect, would also serve the employer's interest in 'efficient and trustworthy workmanship.' "); Blake v. City of Los Angeles, 595 F.2d 1367, 1383 (9th Cir. 1979); Pettway v. American Cast Iron Pipe Co., 494 F.2d 211, 245-46 (5th Cir. 1974); United States v. Bethlehem Steel Corp., 446 F.2d 652, 662 (2d Cir. 1971); Title VIII, e. g., Resident Advisory Board v. Rizzo, 564 F.2d 126, 149 (3d Cir. 1977), and in at least one instance under Title VI, NAACP v. Wilmington Medical Center, 491 F. Supp. 290 (D.Del.1980).

The consideration of alternatives that has occurred in Title VII cases is instructive as to the appropriate standard for challenges under Title VI. Title VII cases typically involve a challenge to a particular selection device, frequently an examination. If the selection device has a disparate racial impact, there is a compelling argument for prohibiting its use, despite its job-relatedness, if another device will also select qualified employees and have a lesser disparate impact. In that context the inquiry is sharply focused upon comparable alternatives other selection procedures or examinations. With Title VI, however, the inquiry could frequently become too open-ended. If, for example, a court were to assess alternative ways of saving funds throughout the administration of a city or even throughout the administration of the health care function, it would seriously risk substituting its own judgment for that of the city's elected officials and appointed specialists. We are skeptical of the capacity and appropriateness of courts to conduct such broad inquiries concerning alternative ways to carry out municipal functions. Once a court is drawn into such a complex inquiry, it will inevitably be assessing the wisdom of competing political and economic alternatives. Moreover, such policy choices would be made without broad public participation and without sufficient assurance that the alternative selected will ultimately provide more of a benefit to the minority population.

In Title VII cases, the inquiry into alternatives is focused on considering alternative means of selecting the number of employees that the employer has decided to hire, for the type of job that the employer has designated. No inquiry is made into whether the employer could reduce the racially disproportionate impact of his selection procedure by considering such broad alternatives as hiring additional employees or changing the method of production. In this case, an inquiry comparable to a Title VII consideration of alternative selection procedures would be consideration of whether closing any other municipal hospitals could have achieved the City's objectives with less disparate racial impact. Even a Title VI inquiry focused primarily upon alternative locations for placement or closing of facilities extends courts beyond their traditional functions. See, e. g., NAACP v. Wilmington Medical Center, supra. We do not foreclose the possibility of a situation where some arrangement, other than the closing of another facility, has such obvious advantages that it must be considered as an alternative to a closing with a significant disproportionate racial impact. However, in this case, we do not believe Title VI requires consideration of alternatives beyond an assessment of all the municipal hospitals in order to select one or more for closing. The City made that assessment, and the appropriateness of their choice of Sydenham for closing was sufficiently demonstrated in court. Cf. West v. Board of Education, 612 F.2d 644 (2d Cir. 1979). The alternatives plaintiffs wish to have considered are more appropriate for examination by administrative, legislative, and other political processes than by the courts.4

Without expressing any views as to the wisdom of closing Sydenham Hospital, we conclude that the plaintiffs have shown no likelihood of success in establishing that its closing would violate Title VI, and the District Court therefore did not err in denying a preliminary injunction.

Wholly apart from whether a Title VI violation might ultimately be found, the Boyd plaintiffs contend that a preliminary injunction is required to maintain the status quo pending the completion of HHS's administrative investigation. Though HHS has never moved for such an injunction on its own behalf,5 the United States as amicus contends that a private plaintiff in a Title VI suit may secure a preliminary injunction pending an administrative investigation. The Government acknowledges that there is no "direct authority" for this claim, but suggests that support by analogy may be drawn from cases upholding the right of Title VII plaintiffs to obtain a preliminary injunction pending investigation of discrimination charges by the Equal Employment Opportunities Commission (EEOC). See, e. g., Garza v. Texas Educational Foundation, 565 F.2d 909 (5th Cir. 1978); Drew v. Liberty Mutual Insurance Co., 480 F.2d 69 (5th Cir. 1973).

We do not find the Title VII analogy persuasive. A plaintiff claiming employment discrimination in violation of Title VII is required to present his claim to the EEOC and afford that agency an opportunity to resolve his grievance before he may bring a civil action of his own in the district court. 42 U.S.C. § 2000e-5(f)(1) (1976). Title VI has no comparable requirement. There is a sound basis for permitting a Title VII plaintiff, in appropriate circumstances, to maintain the status quo pending an administrative proceeding that he is obliged to initiate as a condition of his suit. But no similar basis exists for the Title VI plaintiff who is entitled to sue to prevent discriminatory action without awaiting any administrative proceeding. See Regents of the University of California v. Bakke, 438 U.S. 265, 98 S. Ct. 2733, 57 L. Ed. 2d 750 (1978); cf. Cannon v. University of Chicago, 441 U.S. 677, 99 S. Ct. 1946, 60 L. Ed. 2d 560 (1979). The Title VI plaintiff can secure a preliminary injunction against threatened discrimination by showing a probability of success on his Title VI claim and irreparable injury. He need not await an administrative inquiry, and he should not be entitled to secure a delay simply because an administrative inquiry has not been completed. If the pendency of that inquiry is justification for maintaining the status quo, such a claim must be pursued by the agency itself.

The choices forced upon city administrators by the competition for scarce tax dollars are difficult and frequently unpleasant. There is understandable cause for distress when fiscal realities require the closing of a municipal hospital, and that distress is compounded when the hospital selected for closing is one that serves primarily minority residents of an area woefully lacking in adequate medical resources. Obviously, every effort should be made to assure that the City's available resources are wisely used to mitigate as far as possible the consequences of closing Sydenham Hospital. The City has in good faith planned for the hospital needs of those who now used Sydenham. Those plans, as we have recognized, do not absolutely assure complete provision of alternative service, though the general adequacy of alternatives has been sufficiently shown. We would expect the City to remain sensitive to the concerns expressed during this litigation and to explore all realistic proposals for assuring that those served by Sydenham Hospital have their hospital needs fully met. In that expectation, and because we agree with the District Court's conclusion that entitlement to a preliminary injunction has not been shown, the ruling appealed from is affirmed.

KEARSE, Circuit Judge (concurring in part and dissenting in part):

While I concur in the affirmance of the judgment in Boyd v. Harris, No. 80-6085, for the reasons stated in the majority opinion, I dissent from the affirmance of the denial of a preliminary injunction in Bryan v. Koch, No. 80-7401. I would decide the question left open by the majority, and hold that Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d et seq. (1976), condemns conduct which has a disproportionate racial impact or effect, and not merely conduct motivated by a racially discriminatory purpose. Under an effects test I conclude, like the majority, that plaintiffs made a prima facie showing of a Title VI violation, shifting the burden to the defendants (hereinafter referred to collectively as the "City") to show justification for the decision. I do not agree, however, that the City successfully carried that burden. No one would contest the fact that the City must assign priorities among competing economic demands and evaluate political and economic alternatives. But in my view, when a recipient of federal moneys makes a decision to use those moneys in a way which has disparate racial impact Title VI requires the recipient to show, at the very least, that its decision was the product of a rational decision-making process. The City has made no such showing here as to its decision to close Sydenham.

* My views on the proper test to be applied under Title VI may be stated briefly as follows. Title VI was enacted as part of a sweeping package of remedial measures known collectively as the Civil Rights Act of 1964. Section 601, the basic substantive provision of Title VI, broadly provides, without qualification, that:

No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.

42 U.S.C. § 2000d (1976). It might be argued that this provision, standing alone, is on its face ambiguous with regard to the question of intent vs. effects. But § 601 does not stand alone. Section 602 of the Act imposes on all federal agencies "empowered to extend Federal financial assistance to any program or activity," a duty

. . . to effectuate the provisions of section 2000d of this title with respect to such program or activity by issuing rules, regulations, or orders of general applicability which shall be consistent with achievement of the objectives of the statute authorizing the financial assistance in connection with which the action is taken.

42 U.S.C. § 2000d-1 (1976). Shortly after Title VI was enacted, no fewer than seven federal agencies carried out this mandate "to effectuate the provisions of § 601" by promulgating regulations that applied a broad "disparate impact," or "effects," test.1 These regulations which include the regulations at issue in the instant case, 45 C.F.R. Part 80 reflected the agencies' recognition of Congress's intent that Title VI "would, in short, assure the existing right to equal treatment in the enjoyment of federal funds." 110 Cong.Rec. 1519 (1964) (remarks of Rep. Celler, floor manager of the bill in the House of Representatives; emphasis added). To the same effect, see also id. at 6544 (remarks of Sen. Humphrey); id. at 6561 (remarks of Sen. Kuchel); id. at 7058 (remarks of Sen. Pastore). And see id. at 6543 (remarks of Sen. Humphrey, quoting Presidential message to Congress proposing the legislation):

Simple justice requires that public funds, to which all taxpayers of all races contribute, not be spent in any fashion which encourages, entrenches, subsidizes, or results in racial discrimination.

(Emphasis added.)

In Lau v. Nichols, 414 U.S. 563, 94 S. Ct. 786, 39 L. Ed. 2d 1 (1974), the Supreme Court expressly upheld HEW's2 Title VI regulations, establishing that the standard of liability is impact or effects, not intent:

Discrimination is barred which has that effect even though no purposeful design is present: a recipient "may not . . . utilize criteria or methods of administration which have the effect of subjecting individuals to discrimination" or have "the effect of defeating or substantially impairing accomplishment of the objectives of the program as respect individuals of a particular race, color, or national origin." (45 C.F.R.) § 80.3(b)(2).

414 U.S. at 568, 94 S. Ct. at 789 (emphasis in original).3 Thus these remarks by the legislators, the implementing regulations of the responsible federal agencies, and the one Supreme Court decision on point all reflect the principle as an element of "simple justice" that a recipient of federal funds should not use those funds in a manner which has the effect of excluding persons on the basis of race, color or national origin.

The controlling precedent in this Circuit is to the same effect:

Title VI findings of discrimination may be predicated on disparate impact without proof of unlawful intent.

Board of Education v. Califano, 584 F.2d 576, 589 (2d Cir. 1978), aff'd on other grounds sub nom. Board of Education v. Harris, 444 U.S. 130, 100 S. Ct. 363, 62 L. Ed. 2d 275 (1979). And other circuits have applied the same standard. E. g., Guadalupe Org., Inc. v. Tempe Elem. School Dist., 587 F.2d 1022, 1029 (9th Cir. 1978); Serna v. Portales Municipal Schools, 499 F.2d 1147, 1153-54 (10th Cir. 1974). In fact, with the exception of cases involving school desegregation, it appears that no court has imposed the intent standard in a Title VI case.4

The court below opined that dicta in the concurring opinion of four justices in Regents of the University of California v. Bakke, 438 U.S. 265, 324, 98 S. Ct. 2733, 2766, 57 L. Ed. 2d 750 (1978), and the dissenting opinion of Justice Stewart in Board of Education v. Harris, 444 U.S. 130, 152, 100 S. Ct. 363, 375, 62 L. Ed. 2d 275 (1979), had undermined the authority of Lau v. Nichols, supra. These dicta do not justify the district court's burial of Lau. The majority in Board of Education v. Harris, for example, expressly declined to decide the issue of whether Title VI incorporated an intent test. 444 U.S. at 149, 100 S.Ct. at 374. And most recently, a plurality opinion of the Court in Fullilove v. Klutznik, --- U.S. ----, 100 S. Ct. 2758, 65 L. Ed. 2d 902 (1980), invoked Lau in support of the Court's decision.5

Since the contemporaneous agency regulations and, in my view, the legislative history, clearly support application of an effects test, I believe we are correct to wait for an actual decision from the Supreme Court in a case fairly presenting the issue, before we assume that Lau v. Nichols has gone by the boards. Lau v. Nichols is precedent which clearly requires application of an effects test to the instant case, and I therefore would not be deflected by dicta in cases devoted to the resolution of other weighty issues.

The majority concludes, and I agree, that plaintiffs made a prima facie showing that the closure of Sydenham Hospital would have a disproportionate racial impact. Under the effects test, therefore, the issue on this appeal is whether the City has carried its burden of demonstrating a sufficient nondiscriminatory justification for the closure. See NAACP v. Wilmington Medical Center, 491 F. Supp. 290, 314 (D.Del.1980). Cf. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973) (Title VII); Robinson v. 12 Lofts Realty, Inc., 610 F.2d 1032 (2d Cir. 1979) (Title VIII). Accordingly, this Court must determine whether the decision to close Sydenham was, in some sense, a "proper," or at least an "acceptable" one.

The threshold question, of course, is what may constitute justification for a decision that has prima facie been shown to have a discriminatory impact. Must the City show that it had no alternative to the course chosen? Must it show that it chose the best existing alternative? May it prevail by showing that it considered the existing alternatives and did what it thought best? May it prevail by showing that it considered the existing alternatives and then flipped a coin? I suggest that the most appropriate course is a two-phase evaluation of the purported justification, under which the court would first examine the process by which the decision was reached, and only then, if a rational course has been followed, inquire into the substantive merit of the decision.

Having made a decision which will have a discriminatory impact, a Title VI defendant should be required to show at the very least that it proceeded on a rational basis in its decision-making process. In my view a rational process must include consideration of appropriate alternatives in the light of a factual assessment of the effect of these alternatives. If the defendant fails to show such a rational basis for the decision we should proceed no further in the justification inquiry. Such a defendant cannot be permitted to justify its decision by simple reference to the substantive goals that it sought to advance, because if its process lacks rationality, the defendant will not know either the extent to which its goals are in fact advanced, or the price of their advancement. Ending the inquiry at the point at which it is determined that the defendant has failed to show a rational basis for its decision has several advantages. It will spare the court the task, undertaken by the majority here, of attempting its own evaluation of suggested alternatives whose merits have been inadequately developed in the record. It will avoid the need to speculate as to what the defendant's decision would have been had it actually known all the facts. And it will minimize the risk of judicial usurpation of executive prerogatives inherent in any review of the substance of the decision.

I see no need in the present case to reach the question of the substance of the City's decision because in my view the City failed to show that its decision had a rational basis.

Before turning to the facts, it is important to note the precise posture of this case as it comes to us. The propriety of the district court's denial of preliminary relief turns on whether plaintiffs established, to some degree,6 their probable success on the merits of their claims. At the hearing below, plaintiffs made a prima facie showing that the closure of Sydenham would have a disproportionate racial impact, and the burden therefore shifted to defendants to justify the closure.

Unfortunately, while the district court purported to apply effects test analysis, it apparently did so only as an afterthought. This issue was treated in one summary paragraph at the end of a 50-page opinion, and the absence of any supporting analysis of the facts renders the lower court's conclusion that justification had been established less than persuasive. The bulk of the opinion was devoted to whether effect or intent was the proper test and to the evidence of discriminatory intent. Virtually all of the district court's factual analysis was directed at the latter question. It has therefore been left to this Court to wade through the record and to reconstruct the case along the lines of plaintiffs' consistently asserted, and prima facie valid, Title VI claim. For these reasons, as well as for those stated by the majority, I agree that this is a proper case for relaxation of the standard of review of denials of preliminary relief.

B. The Inadequacy of the Attempted Justification

The underpinnings of the City's defense of justification are the propositions that it needs to close Sydenham in order to save money, and that it has determined that this closing would not unduly affect Sydenham's patients. The deficiencies I find in the City's presentation are that it took an unreasonably narrow view as to the range of possible alternatives, and that it had insufficient information as to the actual effects of the closing on Sydenham patients to make meaningful an evaluation of the available alternatives.

In simplest terms, the question facing the City was whether closing Sydenham was an "appropriate" way to save money. No one questions the City's need to save money. Appropriateness, in this context, therefore turns on what alternative methods of saving money are available. This Court is not in a position precisely to delineate all of the alternatives which the City should have considered, but it is possible and appropriate to settle upon a reasonable area of inquiry. The allocation of City expenditures as among education, sanitation, transportation, fire fighting, police protection, health services, and so forth, is clearly a political decision. I would not venture to say that the City should have considered decreasing expenses in these other areas in order to avoid closing Sydenham.7 I thus agree with the majority that the alternatives to be considered should be confined to those within the health services area, and am willing to assume that our inquiry should be limited to various measures that might have been taken within the Health and Hospitals Corporation ("HHC"). But I cannot subscribe to the majority's assumption that closing one of the other 16 municipal hospitals was the real much less the only alternative to closing Sydenham. There are indications in the record, amounting to little more than statements of the common sense principle, that from the point of view of cost-efficiency the closing of a whole hospital will generally save more money than would a comparable reduction in beds effected by elimination of single beds, wards or wings within one or more hospitals. This is so because many overhead expenses that are difficult to reduce when the hospital merely shrinks, can be eliminated entirely when the hospital is closed. If cost-efficiency were the sole criterion, then the only alternatives to closure of Sydenham would probably have been closure of other hospitals. But it was not in fact the sole criterion: this was made clear, for example, by the Mayor's Report, which purported to apply a much more complex analysis to the problem of eliminating excess capacity. One of the considerations applied in this analysis was patient access to comparable alternative facilities. The report recommended that some hospitals not be closed solely on the basis of this consideration.

The complexity of the problem is also suggested by the recommendations which the Mayor's Report actually made. These included a wide range of cost-cutting measures, such as elimination of individual beds and of larger units within hospitals, as well as closure of two whole hospitals. It is therefore clear that the choices facing the City were in fact quite complex. There is no reason to assume a simple but artificial set of alternatives simply for purposes of evaluating plaintiffs' Title VI claim. But this is what the majority does when it compares the closing of Sydenham only to the possibility of closure of some other hospital.

The plaintiffs have contended that there were perfectly sound money-saving alternatives to the closing of Sydenham, involving mergers of municipal hospitals plus the regionalization of services to promote other efficiencies. They presented testimony of a former HHC official to show the feasibility of such alternatives. The City sought to show that the possibility of savings resulting from mergers was speculative; but it remains true that the City's position that this was not a feasible alternative was developed only during the course of this lawsuit. In the course of deciding to close Sydenham the City had not evaluated these possibilities, despite the fact that a city official recommended the study of possible savings from hospital consolidations, and despite evidence at trial that a study of the alternatives suggested by plaintiffs could have been completed in just three weeks' time. In my view, the failure of the City to consider significant possible alternatives is a serious defect in its decision-making process.

Even had the City considered all of the alternatives for health care savings, however, I would not conclude that it had justified its decision to close Sydenham. This is so because throughout the process, the City's decisionmakers focused so closely on economic considerations that they ignored reality as to the impact that the various alternatives would likely have on Sydenham's patient population.

In fiscal year 1979, Sydenham admitted 3711 inpatients and had an average occupancy of 93 inpatients. The hospital currently has 107 certified beds. It is therefore a relatively small part of HHC's overall operations, and its occupancy rate has been somewhat lower than the 90% figure generally considered optimal. Sydenham Hospital has other problems as well, including an older physical plant and a low Medicaid reimbursement rate, which render its operation relatively cost-inefficient. In economic terms, therefore, the City was clearly justified in considering Sydenham as a candidate for shrinkage or closure.

Nevertheless, certain characteristics of Sydenham's clientele tend to distinguish them from the patient populations of other municipal hospitals. Primary among these distinctions are the fact that 63% (in fiscal year 1979; the proportion has been steadily increasing) of Sydenham's inpatients are admitted through the emergency room, and the fact that at least one third8 of the persons presenting themselves to Sydenham's emergency room do not have medical insurance (i. e., Medicaid, Medicare, Blue Cross/Blue Shield or other coverage) at the time they come in. These two facts become critically important when we consider the City's analysis of what will happen to Sydenham's patients if that hospital closes. The City introduced the results of a hypothetical study that allocated Sydenham's patients to surrounding hospitals which, according to the statistics relied upon, consistently have unoccupied beds. The study placed particular reliance on four hospitals: St. Luke's, Presbyterian and Mt. Sinai, which are voluntary hospitals, and Harlem, which is municipal. Plaintiffs, however, presented persuasive evidence that the three voluntary hospitals have a practice of not admitting uninsured patients i. e., patients who cannot produce proof of insurance at the time of admission. The City never asked the voluntaries about their policy in this regard. Hence it is not in a position to rebut plaintiffs' showing. The City's study purported not to allocate uninsured patients to the voluntary hospitals, but since the study assumed a much lower proportion of uninsured patients than appears to exist in fact, the study necessarily (even if unintentionally) did make such allocations. The study therefore does not adequately account for Sydenham's uninsured patients. Moreover, there is no basis in this record for determining the actual scope of the problem.

The City raises two arguments in this regard, neither of which is persuasive. First, it points to New York Pub. Health Law § 2805-b(1) (McKinney 1977), which provides that

Every general hospital shall admit any person who is in need of immediate hospitalization with all convenient speed and shall not before admission question the patient or any member of his or her family concerning insurance, credit or payment of charges . . .

The City argues that under this section, the voluntary hospitals would be required to admit some uninsured patients. It is far from clear, however, that this provision would aid most of the Sydenham patients, because only the true emergency patients are provided for by this section. While it is a fact that most of Sydenham's patients are admitted from its emergency room, a relatively small percentage of these are true emergency patients.9 There is no reason to believe that the voluntary hospitals would admit the others.

The City also argues that even when provision for Sydenham's uninsured patients is evaluated on the basis of plaintiffs' statistics, there will be only an insignificant number of uninsured patients seeking admission as inpatients. The City argues in its reply brief that if there were 3767 admissions to Sydenham in calendar year 1979, and 63% (or 2373) were admitted through the emergency room, then we can conclude that only 33.6% of these (the proportion of emergency room visitors who are uninsured, equalling 797 patients, or 2.18 per day) will be uninsured patients seeking admission at the surrounding hospitals, after Sydenham closes. Thus, according to this brief, "The number of uninsured patients is so small (i. e., only about 2 inpatient admissions per day) that the record below amply demonstrates that they would be cared for at Harlem Hospital . . . ." (City Reply Brief at 2). One fundamental flaw in this argument is its apparent disregard of the fact that, once admitted, the average Sydenham patient remains hospitalized for about 10 days. Thus plaintiffs point out that on any given day Sydenham has in its beds 23-28 uninsured patients who would need to be cared for elsewhere if Sydenham closed. The City's prediction that Harlem Hospital would care for two uninsured Sydenham patients per day is of little value to the vast majority.

There are other problems with the City's analysis of the closure's impact. For example, that analysis was based on hospital occupancy figures for past years. Plaintiffs introduced a variety of evidence apparently uncontroverted that more recently the occupancy rates in Harlem area hospitals had increased, so that frequently all of the beds at Harlem Hospital,10 Presbyterian and St. Luke's are filled, and that apparently all of these hospitals are now constantly operating at close to maximum capacity. The district court apparently credited this evidence. See 490 F. Supp. 291 at n.14. There is no indication, however, that the City's decisionmakers considered these new developments.

Another possible problem involves Sydenham's Medicare and Medicaid inpatients. The voluntary hospitals in Sydenham's area maintain a lower percentage of inpatients in these categories than does Sydenham or the other municipal hospitals. Since the Medicare and Medicaid payments do not cover the entire cost of treatment, the voluntary hospitals have a clear disincentive to take a greater number of patients in these categories.11 The City apparently did not consider this factor in assessing the impact of Sydenham's closure.

The foregoing discussion, while hardly exhaustive, is sufficient to demonstrate that the City did not give adequate consideration to the actual impact of closing Sydenham. The City's patient redistribution studies were nothing more than hypothetical constructs. This is well illustrated by the testimony of the City's expert, referring to one aspect of the study:

It was a device and it doesn't represent either a reality, which none of this construct does, nor does it represent any real percentage of Medicaid days which we either knew or believed would go to any one of these hospitals. It's just a construct to show an order of magnitude figure on the bottom line.

(Transcript of Hearing at 1332.) Plaintiffs have convincingly demonstrated that the City's constructs ignored major practical problems which Sydenham's patients would encounter after a closure. The City did not try to find out what policies the voluntary hospitals actually follow, nor did it adequately analyze the actual availability of alternative hospital facilities. It simply made little or no effort to determine what would actually happen to Sydenham's patients upon closure of that hospital.

I an unable to find justification on the basis of such untested hypotheses or on assumptions which are untried and quite possibly untrue. I believe plaintiffs were entitled to a preliminary injunction against the closure.