The opinion of the court was delivered by: SOFAER
These actions all seek to prevent Mayor Edward I. Koch, and others, from implementing a plan to close certain municipal hospitals in the City of New York. At issue now is the closure of Sydenham Hospital, since the City has yet to decide whether other proposed closings should be implemented.
The defendant Health and Hospitals Corporation ("HHC") operates the City's municipal hospital system, which includes 13 acute care hospitals and four long-term care facilities for the treatment of the chronically ill and infirm. New York City's municipal hospital system is by far the largest system of its kind in the country.
The hospitals in the municipal system provide health care to anyone who presents himself for treatment, without regard to ability to pay. Consequently, the budget of HHC constitutes over 10% of the total expense budget of the City of New York for fiscal year 1980 about.$ 1.2 billion. In fiscal year 1980, the City of New York will contribute over $ 500,000,000 of tax funds to subsidize the operations of the hospital system. City Defendants' Trial Mem., p. 20.
Sydenham Hospital, located at 565 Manhattan Avenue near West 124th Street, is a provider of general, inpatient hospital care.
It is the smallest of New York's 13 acute care municipal hospitals. In fiscal year 1979, Sydenham admitted 3711 inpatients (1.5% of HHC's total), treated 25,690 emergency room patients (1.8% of HHC's total), and housed an average of 93 inpatients per day.
Sydenham serves primarily the population of the Central Harlem and West Harlem communities east of Morningside Heights. Most of the residents of these communities are black.
Plaintiffs contend that closing Sydenham would violate the equal protection clause of the Fourteenth Amendment to the United States Constitution, and Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d et seq.
They represent three separate groups, two of which seek class certification.
On February 8, 1980, plaintiffs in the Bryan and District Council 37 cases moved to enjoin the City defendants from closing Sydenham pending a full trial on the merits, or alternatively until defendants guaranteed inpatient and emergency services to the population served by Sydenham.
A lengthy hearing was held on this motion. Numerous witnesses testified. Affidavits supplemented the live testimony. And a massive record of exhibits has been compiled. The hearing ended on April 17, 1980. Post-hearing submissions continue to be made as of May 15, 1980.
The Mayor has stated his intention to order Sydenham closed on May 16, 1980. The schedule that his decision compelled has been exacting. Under normal circumstances, a temporary restraining order would have been justified to permit the preparation and filing of detailed findings and conclusions.
But this is no ordinary case. It appears, rather, to be an effort by plaintiffs to use the federal courts as a last resort for delaying if not preventing the implementation by elected officials of a painful but purely political decision. Under these circumstances, to delay the closing of Sydenham for any period particularly for the decision-making convenience of this court would serve to undermine the authority and governing capacity of the City's responsible officials.
I. Standard for Preliminary Relief
The Second Circuit has made clear that preliminary relief is appropriate only where the party seeking it shows possible irreparable injury. In addition, the moving party must establish:
either (1) probable success on the merits or (2) sufficiently serious questions going to the merits to make them a fair ground for litigation and a balance of hardships tipping decidedly in the movant's favor.
E. g., Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206-07 (2d Cir. 1979).
Defendants urge that relief be denied simply because no possible irreparable injury has been shown. This is so, if for no other reason than the fact that plaintiff's alternative request for preliminary relief guaranteed access to inpatient and emergency services for Sydenham patients without unreasonable burdens has been demonstrated. See note 6, supra. In this case, however, the lack of merit in plaintiffs' contentions even more clearly mandates denying preliminary relief than the absence of irreparable injury.
II. Likelihood of Success on the Merits
Plaintiffs present several claims. All relate ultimately to the contention that defendants have violated the rights of minorities in attempting to close Sydenham. Some clearly lack merit, such as the assertion by plaintiffs in Boyd that Sydenham must be kept open until the Department of Health and Human Services ("HHS") (formerly Health, Education and Welfare, "HEW") completes its investigation as to the City's compliance with Title VI.
Most are complicated claims, however, and so novel that pains have been taken to compile a record on issues which reviewing courts may see fit to address. In particular, this case requires federal courts to determine when the reduction of services to minorities by recipients of federal funds violates the Constitution and laws of the United States. The answers we give to this question are important in present times, when reductions in government services have become increasingly common, particularly in areas heavily populated by minorities.
A. Claimed Violation of Equal Protection
The legal standard that governs relief under the equal protection clause of the Fourteenth Amendment requires proof that the challenged law or conduct "ultimately be traced to a racially discriminatory purpose." Washington v. Davis, 426 U.S. 229, 240, 96 S. Ct. 2040, 2048, 48 L. Ed. 2d 597 (1976). Discriminatory impact is insufficient in itself to support a conclusion of discriminatory purpose. Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 270, 97 S. Ct. 555, 566, 50 L. Ed. 2d 450 (1977); Washington v. Davis, supra, 426 U.S. at 242, 96 S. Ct. at 2048. Furthermore, although the foreseeability of a racially adverse impact is still a proper consideration in determining discriminatory purpose,
a foreseeable adverse impact in itself is insufficient to prove discriminatory purpose. Personnel Administrator of Mass. v. Feeney, 442 U.S. 256, 279 n. 25, 99 S. Ct. 2282, 2296, 60 L. Ed. 2d 870 (1979); Columbus Board of Education v. Penick, 443 U.S. 449, 99 S. Ct. 2941, 61 L. Ed. 2d 666 (1979); Dayton Board of Education v. Brinkman, 443 U.S. 526, 99 S. Ct. 2971, 61 L. Ed. 2d 720 (1979).
Plaintiffs properly argue that these new standards, although demanding, permit parties to prove discriminatory purpose by inference from objective evidence. "Determining whether invidious discriminatory purpose was a motivating factor (requires) a sensitive inquiry into such circumstantial and direct evidence of intent as may be available." Village of Arlington Heights v. Metropolitan Housing Development Corp., supra, 429 U.S. at 266, 97 S. Ct. at 564. "Proof of discriminatory intent must necessarily usually rely on objective factors," Personnel Administrator of Mass. v. Feeney, supra, 442 U.S. at 279 n. 24, 99 S. Ct. at 2296 n. 24 such as the fact "that the law (or practice) bears more heavily on one race than another," Washington v. Davis, supra, 426 U.S. at 241-42, 96 S. Ct. at 2049, the fact that the "actions (undertaken had) foreseeable and anticipated disparate impact . . . " Columbus Board of Education v. Penick, supra, 443 U.S. at 464, 99 S. Ct. at 2950, or the presence of proof of segregative intent based upon the particular history of the challenged decision, Village of Arlington Heights v. Metropolitan Housing Corp., supra, 429 U.S. at 266-68, 97 S. Ct. at 563-65.
Although these rules allow plaintiffs to attempt to prove discriminatory purpose without any direct evidence of such intention on the part of any, let alone all, relevant decisionmakers, the Court requires that the ultimate issue racial animus be kept in view. Discriminatory purpose "implies that the decisionmaker . . . selected or reaffirmed a particular course of action at least in part "because of,' not merely "in spite of,' its adverse effects upon an identifiable group." Personnel Administrator of Mass. v. Feeney, supra, 442 U.S. at 279, 99 S. Ct. at 2296.
Plaintiffs claim they can prove discriminatory purpose in this case from evidence of (1) foreseeable adverse impact upon blacks and Hispanics by the closing of Sydenham; (2) the City's failure in deciding to close the hospital to adhere to substantive norms of health and fiscal planning; (3) the City's inconsistent application of its own criteria for selecting hospitals for closure, bed reductions, or mergers; (4) the City's reliance on erroneous data and arguments, even after "official criticism" by other health planning agencies; (5) the failure to consider feasible and less onerous alternatives; and (6) the City's alleged "nine months of evasion of a census of patients in the municipal hospitals," requested by HHS's Office of Civil Rights. See Plaintiffs' Post Hearing Mem., pp. 227-30.
1. Foreseeable Disparate Impact
Plaintiffs claim that data collected on the racial composition of persons who use Sydenham and other municipal hospitals demonstrate that closing Sydenham will have a disproportionately adverse impact upon minorities, and that the City's hospital plan as a whole would, if implemented, have a disparate impact upon minorities. In making this claim plaintiffs proved a disparate impact of radically different meaning and significance than the disparate impact the Supreme Court has relied upon in jury-selection and other cases to prove discriminatory purpose.
Plaintiff's expert, a sociologist named Richard Faust, relied on several sets of data concerning proposed hospital or bed closings. Disputes as to the relative accuracy of these data sets have raged needlessly since this case was filed. No one disputes that approximately two-thirds of the patients who use the City's municipal hospitals are black or Hispanic. (Exh. 12, 13) Furthermore, virtually all the patients who use Sydenham are minority. Faust took these facts, among others, and designated the beds in all municipal hospitals, including Sydenham, in accordance with the racial proportion of patients. Thus, if a hospital had 1000 beds and 80% minority patients, 800 of its beds were designated minority beds. Faust then calculated the total number of minority beds that were scheduled to close in the Mayor's Plan, and compared that number to the proportion of minority beds that would have been "expected" to be reduced if the percentage of minority beds to be reduced were the same as the proportion of minority beds in the municipal hospital system. He found, regardless of what data set he used, that more minority beds were proposed for reduction by the Plan than would have been expected had reductions been in the same proportion as minority beds in the municipal hospital system. He then treated this difference as a disparity or, as plaintiffs contend, a disparate impact and proceeded to calculate the likelihood that the disparities found could have been the result of random distribution. In every case analyzed he found disparities of at least two standard deviations from the expected distribution, and in several cases he found disparities in excess of ten standard deviations. (Exh. 57e-g, 91)
Defendants' expert, a statistician named Dr. William B. Fairley, made clear how little utility plaintiffs' data possess. To utilize the data presented by plaintiffs on the issue of discriminatory purpose, Dr. Fairley explained, one would have to assume that the decisionmaker whose motivations were under scrutiny made individual, independent decisions to reduce beds in the system or in any hospital studied. A finding of discrimination may theoretically be inferred from a highly improbable selection of minority beds for reduction. But to draw such an inference the selection of beds must be as independent, one from the other, as the selection of jurors in Castaneda v. Partida, 430 U.S. 482, 97 S. Ct. 1272, 51 L. Ed. 2d 498 (1977), Swain v. Alabama, 380 U.S. 202, 85 S. Ct. 824, 13 L. Ed. 2d 759 (1965), or as the selection of teachers for assignment in Board of Education v. Harris, 444 U.S. 130, 100 S. Ct. 363, 62 L. Ed. 2d 275 (1979). See Finkelstein, The Application of Statistical Decision Theory to the Jury Discrimination Cases, 80 Harv.L.Rev. 338, 374 (1966). The relevant statistical technique is the binomial model, in which one of two choices is made over and over again, creating the basis for testing the probability of the distribution of some total number of independent, individual choices. See, e.g., F. Mosteller, R. Rourke & G. Thomas, Probability with Statistical Applications 132-33 (2d ed. 1973) (Exh. O).
The Mayor and his administrators in this case obviously made no decisions respecting individual bed reductions. They proposed hospital closings, and multiple bed reductions in specific hospitals. Therefore, any effort to infer motive from the effects of such decisions on individual beds is untenable.
Mr. Faust conceded his analysis had no value in ascertaining a decisionmaker's purpose. He and plaintiffs contend, however, that the analysis is a useful demonstration of the degree to which minorities would be affected by the Mayor's Plan, relative to whites. This would appear, in some crude sense, to be true, and some authority exists for utilizing the binomial model, even though independent decisions could not be assumed, "as a baseline for comparative purposes. . . ." F. Mosteller et al., supra, at 133-34. But the dangers in so simplistic an approach are substantial, as the circumstances of this case illustrate.
When properly analyzed, the data plaintiff advanced to prove disparate impact in fact tend to negate any intent to discriminate. Dr. Fairley illustrated this with three different models of statistical analysis: Fisher's Exact Test, the Mann-Whitney Test, and the Permutation Test. These models tested with differing degrees of sophistication whether the limited number of decisions proposed by the City concerning hospital closings or major bed reductions are consistent with results that could have been obtained by random selection. No effort has been made to controvert Dr. Fairley's convincing analyses, all of which showed that the proposals in the Mayor's Plan could reasonably have been reached by random selection, thereby tending to negate the hypothesis of discriminatory intent. See F. Mosteller & R. Rourke, Sturdy Statistics 194, 56, 268-69, 12-14 (1973). (Tr. 934-86, 1485-87)
Furthermore, as Dr. Fairley explained, any effort to find significance in the limited set of data in this case is suspect. Few choices were available to the Mayor's Task Force in this case; thus very large apparent differences in the effects of their decisions were relatively insignificant as proof of their intent. (Still fewer choices were available to the Mayor.) By contrast, for example, when thousands of jurors are being individually selected, even apparently small differences may produce significant results. Moreover, the nature of the decisions involved in this case is such that testing them against an hypothesis of random distribution is peculiarly inapt. In a typical jury selection case, the decisionmaker is presented with few variables, all relatively simple to apply (such as ascertaining a prospective juror's residence). One would expect, therefore, that the consequence of juror selections will be a pattern of decisions consistent with a random selection process. Decisions to close hospitals, by contrast, or to reduce the number of beds in hospitals, are not the sort of judgments one would necessarily expect to find exercised in a manner consistent with randomness. Rather, one would expect that such judgments are made on the basis of a process far more complex, and involving many more variables. In this context, a proper test of statistical significance based on a random distribution hypothesis may produce "interesting" results, as Fairley put it, but of significantly less value than the results of a less complex set of decisions. (Tr. 1501)
A further danger in the inapposite use of binomial analysis, even as a crude measure of impact on minorities, is that the impact observed may affirmatively mislead as to the decisionmaker's motive. The decision to close Sydenham is only one of literally thousands of individual budgetary determinations by which the City's available resources are distributed. Any meaningful inquiry to determine motivation in such a context would require an evaluation of whether City funds are being measurably reallocated away from services to minorities. The proposed Sydenham closing may well be part of an overall reduction in city expenditures, for example, that falls more heavily on services to whites, and allocates more funds proportionately (even if less absolutely) to services from which minorities predominantly benefit.
Finally, the Court's decision in Personnel Administrator of Mass. v. Feeney, supra, is instructive in evaluating the sort of data plaintiffs present. The veterans preference law challenged in that case had the effect of excluding almost all women from eligibility for certain jobs. This disproportionate impact, however great, could not be deemed probative of improper purpose, the Court concluded, because a believable and historically accepted explanation for the statute was available the policy of preferring veterans. The result of disadvantaging women may have been foreseeable, but this proof was inadequate in light of an altogether believable, non-discriminatory motivation. Here, too, non-discriminatory motives plainly exist fiscal necessity and health planning among others. Unless these motivations are refuted by other indicia of discriminatory purpose, the alleged disparate impact is entitled to little, if any, weight.
2. Other Evidence of Discriminatory Purpose
Plaintiffs' other arguments concerning discriminatory purpose are all related to the manner in which the City has planned and implemented the closing of Sydenham. The searching evaluation which plaintiffs have made of the City's plans, projections and conduct has revealed imperfections, errors, and perhaps excessive combativeness on the part of the City's representatives. But such shortcomings are inevitable, given the limitations under which City planners work, the need for swift judgments on the basis of imperfect data, and the enormous political pressures that have been brought to bear upon the City to prevent its plan from taking effect.
The context in which the City developed and promulgated the Mayor's Plan is uncontroverted. A grave fiscal crisis was widely recognized to exist in New York City in 1975. The City's budget was estimated to be from $ 6 billion to $ 9 billion in overall deficit, approximately $ 2 billion attributable to that year. Further deficits were anticipated. The City was threatened with bankruptcy when it was denied access to normal channels of credit. (Tr. 1352-53)
Severe fiscal constraints followed. As the City's budget was cut, thousands of municipal employees were dismissed. The City became obliged to plan for further reductions, with a view to achieving a balanced budget by fiscal 1981. The Emergency Financial Control Board was created to review the City's plans and to monitor its compliance with them.
At the center of the City's efforts to plan for its necessary retrenchment has been its Office of Management and Budget ("OMB"). Its Deputy Director, Paul Dickstein, testified that, despite substantial cuts in other service areas, HHC had suffered relatively inconsequential reductions in support as of 1978. In addition, by November 1978 it became clear that HHC had overstated its revenues, and would require $ 60-$ 80 million in additional tax levy funds. This came on top of an already projected City budget gap of $ 400 million for fiscal 1980. Despite this, in its January 1979 Plan to Eliminate the Budget Gap, the City proposed to eliminate $ 250 million in spending without any cuts at HHC. But the difficulty of cutting further from the relatively small amount of discretionary spending in the City budget, and the severe cuts already imposed upon other essential services, such as the police and fire forces and the Board of Education, led OMB to establish a program under which HHC would ...