Roselli v. Affleck, 373 F. Supp. 36 (D.R.I. 1974)
February 25, 1974
John J. AFFLECK, Individually and in his capacity as Director of the Rhode Island Department of Social and Rehabilitative Services.
United States District Court, D. Rhode Island.
*37 *38 Cary J. Coen, and Jay C. Lipner, of Rhode Island Legal Services, Inc., Providence, R. I., for plaintiffs.
W. Slater Allen, Jr., Asst. Atty. Gen., Providence, R. I., for defendant.
PETTINE, Chief Judge.
The named plaintiffs bring this action on their own behalf and, pursuant to Rule 23, Federal Rules of Civil Procedure, on behalf of all other Rhode Island recipients of financial assistance under the Aid to Families with Dependent Children Program, "AFDC", who are residing in and obligated to pay for housing other than that financed and operated by Public Housing Authorities created pursuant to General Laws of Rhode Island, section 45-21-1 et seq.
On November 1, 1973 the defendant instituted a so-called "flat grant" system providing for a consolidated payment to "AFDC" recipients whereby all family units of the same numerical size receive precisely the same assistance payment. Within this consolidation there is an allotment for shelter which the plaintiffs contend is in conflict with the equal protection clause of the Fourteenth Amendment of the Constitution of the United States and contravenes the Social Security Act of 1935, as amended 42 U.S.C. sec. 602(a) and the regulations promulgated thereunder by the United States Department of Health, Education and Welfare, 45 C.F.R. 233.20(a) (1), 233.20(a) (3) (ii), 233.20(a) (3) (iii) (c). Specifically the plaintiffs make a four pronged attack contending that the defendant
1) failed to compute independently those recipient units living in public housing from those living in private housing in computing the average of shelter costs;
2) at the time of averaging, in keeping with its policy in regard to assistance to "non-square" or pro-rated families, irrebuttably presumed that the non-recipient household member was able and did in fact contribute a pro-rated share to defray the living expenses of the *39 "AFDC" recipients and thus included said amount in computing the standard;
3) failed to average or even survey actual rents but instead used in the computation the rents previously budgeted by the state to the recipients; and
4) failed to account for the inflationary increase in rents from the date of the averaging of shelter costs in November 1972, a date when rent increases were tightly regulated under the Economic Stabilization Act of 1970, as amended, P.L. 91-379, 84 Stat. 799 and the regulations issued thereunder by the Price Commission, 6 C.F.R. sec. 501 et seq. to the time of implementation in November 1973.
The plaintiffs argue that these failures, which either singly or collectively, violate the federal regulations cited supra, the Constitution and decisional law established in Rosado v. Wyman, 397 U.S. 397, 419, 90 S. Ct. 1207, 25 L. Ed. 2d 442 (1970),invidiously discriminate against the plaintiff class on a basis not rationally related to the legitimate purposes of the flat grant system; impinge on their right to dwell where they wish without a compelling governmental interest being served; and illegally diminish the content of the standard of need for shelter in that the standard, as established by the state prior to the institution of the flat grant system, was based on actual need; and that in its execution the state often did not meet actual need in violation of its own policy.
All this coalesces into the ultimate complaint that the shelter portion of the "flat grant" is illegal.
The plaintiffs have alleged a colorable constitutional claim. However, the parties have agreed to defer the consititutional issue and agree that the court proceed to hear the pendent statutory claim. The matter thus fits the pattern I followed in Rhode Island Fair Welfare Rights Organization, et al. v. John J. Affleck, CA 4818 (D.R.I. August 7, 1973).
"Although plaintiffs' constitutional claim arguably involves the constitutionality of a state-wide regulation and thus would trigger the requirement of a three-judge court under 28 U.S.C. 2281, 2284, the parties agreed that adjudication of the constitutional issue be deferred and that this Court proceed to hear the pendent statutory claims. Such a procedure would be consonant with the policies of Rosado v. Wyman, 397 U.S. 397 [90 S. Ct. 1207] (1970), that statutory claims be decided before constitutional claims and that non-constitutional claims be heard by a single judge. See also Rhode Island Fair Welfare Rights Org. v. Department of Social and Rehabilitative Services, 329 F. Supp. 860 (D.R.I.1970). While I find no authority from the First Circuit Court of Appeals on this point, I am in agreement with Judge Newman in Connecticut Union of Welfare Employees v. White, 55 F.R.D. 481 (D.Conn.1972) that nothing in 28 U.S.C. sec. 2281 prevents a single judge from hearing pendent statutory claims prior to convening a three-judge court. Savings in judicial time and expediting of the action recommend this procedure." id. at pp. 4-5.
Jurisdiction also exists under 28 U.S.C. sec. 1343(4). In Giguere et al. v. Affleck, 370 F. Supp. 154, at pp. 157-158 (D.R.I.1974) this court stated:
"While other courts may differ on this question, see e. g. Wynn v. Indiana State Department of Public Welfare, 316 F. Supp. 324 (N.D.Ind.1970) I find compelling the reasoning of the Fifth Circuit that an action alleging that under color of state law certain federal statutes or regulations are being violated may state a claim under 42 U.S.C. sec. 1983 even absent an allegation of constitutional invalidity. Gomez v. Florida State Employment Service, 417 F.2d 569 (5th Cir. 1969) [at p. 569] . . . See also Bass v. Rockefeller, 331 F. Supp. 945, 949 N. 5 (S.D.N.Y.1971); Note, Federal Jurisdiction Over Challenges to State Welfare *40 Programs, 72 Colum.L.Rev. 1404, 1417-21 (1972)."
The matter has been certified as a class action under Fed.R.Civ.P. 23(b) (2).
The "Flat Grant"
The State of Rhode Island employed the American Data Systems, Inc. to develop the flat grant. In doing so they selected at random 4265 active "AFDC" cases from the approximately 13,000-14,000 total as of November 1, 1972. They classified this selected group into plan or "family" sizes of 1 through 10 and then surveyed each to ascertain the amount the state budgeted for each need item to be included in the flat grant. Shelter needs were included as budgeted for November 1972. No survey was made to ascertain the actual cost of shelter being paid by the recipients, nor was any attempt made to determine the number of recipients paying rent different from the sum budgeted to them; nor did the survey account for actual rent or budgeted rent increases occurring between the date of the survey and November 1973, the date the flat grant was implemented.
The net of these computations concerning the cost of purchasing shelter are best set forth in the following chart.
Approximately 90% or more of the total of all these columns which is paid to the recipient perhaps represents the rent allotment. See N. 2 & 3, supra.
The "AFDC" program is based on a scheme of cooperative federalism and is largely funded by the Federal Government on a matching fund basis and administered by the states. States are not required to participate in the program, *41 but those which do participate and take advantage of the available federal funds must comply with the terms of the applicable federal legislation, 42 U.S.C. sec. 601 et seq., and the regulations promulgated thereunder by the United States Department of Health, Education, and Welfare, 45 C.F.R. sec. 201 et seq. King v. Smith, 392 U.S. 309, 316, 88 S. Ct. 2128, 20 L. Ed. 2d 1118 (1967); Rosado v. Wyman, 397 U.S. 397, 408, 90 S. Ct. 1207, 25 L. Ed. 2d 442 (1970). States have traditionally been at liberty to pay as little or as much as they choose. Rosado v. Wyman, supra, noted that
"First, it is necessary to establish a `standard of need,' a yardstick for measuring who is eligible for public assistance. Second, it must be decided how much assistance will be given, that is, what `level of benefits' will be paid. On both scores Congress has always left to the States a great deal of discretion. King v. Smith, 392 U.S., at 318 [88 S.Ct., at 2133-2134.] Thus, some States include in their `standard of need' items that others do not take into account. Diversity also exists with respect to the level of benefits in fact paid." id. 408, 90 S. Ct. 1216.
Approving the concept of averaging assistance payments in order to establish a flat grant, the Court in Rosado enunciated, in accordance with 42 U.S.C. sec. 602(a) (23), the basic guidelines to be followed in the conversion procedure. 42 U.S.C. sec. 602(a) (23) provides:
"[The States shall] provide that by July 1, 1969, the amounts used by the State to determine the needs of individuals will have been adjusted to reflect fully changes in living costs since such amounts were established, and any maximums that the State imposes on the amount of aid paid to families will have been proportionately adjusted."
Despite the almost totally nonexistent legislative history behind this statute, the Court ascribed two distinct purposes to 42 U.S.C. sec. 602 (a) (23):
"First, to require States to face up realistically to the magnitude of the public assistance requirement and lay bare the extent to which their programs fall short of fulfilling actual need; second, to prod the States to apportion their payments on a more equitable basis. Consistent with this interpretation of § 402(a) (23), a State may, after recomputing its standard of need, pare down payments to accomodate budgetary realities by reducing the percent of benefits paid or switching to a percent reduction system, but it may not obscure the actual standard of need . . .
It has the effect of requiring the States to recognize and accept the responsibility for those additional individuals whose income falls short of the standard of need as computed in light of economic realities and to place them among those eligible for the care and training provisions. Secondly, while it leaves the States free to effect downward adjustments in the level of benefits paid, it accomplishes within that framework the goal, however modest, of forcing a State to accept the political consequence of such a cutback and bringing to light the true extent to which actual assistance falls short of the minimum acceptable." id. 397 U.S. 412-413, 90 S. Ct. 1218. (emphasis added)
The Court further stated at p. 419, 90 S.Ct. at p. 1221.
"Section 402(a) (23) invalidates any state program that substantially alters the content of the standard of need in such a way that it is less than it was prior to the enactment of § 402(a) (23), . . . We do not, of course, hold that New York may not consistently with the federal statutes, consolidate items on the basis of statistical averages. Obviously such averaging may affect some families adversely and benefit others. Moreover, it is conceivable that the net payout, assuming no change in the level of benefits, may be somewhat less under a streamlined program. Providing all *42 factors in the old equation are accounted for and fairly priced and providing the consolidation on a statistical basis reflects a fair averaging, a State may, of course, consistently with § 402(a) (23) redefine its method for determining need."
Since much of this controversy centers around whether the State of Rhode Island altered and/or obscured its standard of need through its conversion to a new system for aid to "AFDC" recipients, it is imperative the Rosado teachings be carefully scrutinized in juxtaposition with the Rhode Island articulated "standard of need" from July 1, 1969 to October 31, 1973.
From the foregoing language we see that Rosado interprets 42 U.S.C. sec. 602(a) (23) to require the states to update their standard of need as of July 1, 1969 and to take no subsequent action to lower or obscure the updated determination. We further learn that the standard of need is to be considered separate from the level of payment and is determined upon the basis of what the state defines as need. New Jersey Welfare Rights Organization v. Cahill, 349 F. Supp. 501 (D.N.J.1972), aff'd 483 F.2d 723 (3d Cir. 1973). In defining the "standard of need" for rent prior to the inception of the new "flat grant" regulations, Rhode Island chose not to state a specific dollar amount. It established the "standard of need" and the amount of payment upon actual need subject only to the limitation of reasonableness. That is, it paid whatever was needed to meet the recipients' rent requirement for the shelter being occupied. In substance though the "standard of need" and the actual payment may be different Rhode Island chose to make them the same. To state it another way, in Rhode Island the "standard of need", actual need, and payment level are one and the same thing.
To put it simply as of July 1, 1969 the rent "standard of need" established by Rhode Island was on an "as need basis". This being the level which cannot be changed or obscured the question to be answered is whether or not on November 1, 1973 when the flat grant was instituted it did indeed reflect at least the July 1, 1969 policy. I find it did not. Rhode Island failed in complying with the Rosado mandate that, ". . . all factors in the old equation *43 [be] accounted for and fairly priced and [that] the consolidation [be] on a statistical basis reflect[ing] a fair averaging . . ."
In response Rhode Island argues that this means it is only required to pay the same dollar amount it was paying as of July 1, 1969. With all due respect, it appears to me such a position, however sincere, is untenable and ingenious argumentation which cannot be accepted.
Certainly the contention would be accurate if the state had in 1969 computed a specific monetary figure as its standard of need for rent. It did not do this. It paid "actual need" as its "standard of need". Therefore, the payment level necessary to continue to meet the articulated standard becomes a flexible figure varying with need levels. In this process to establish an inflexible figure which falls short of "actual need" undermines the stated function of 42 U.S.C. 602 (a) (23) as interpreted in Rosado.
Having laid the basic foundation for analysis of plaintiffs' cause of action, I shall now consider seriatim the legal merits of each of the plaintiffs' allegations.
Averaging Public Housing Rents with Private Housing Rents
Plaintiffs challenge the legality of averaging public housing rents with private housing rents in computing average rental needs.
Rents in publicly subsidized housing units are substantially lower than rents in the private housing market because of large governmental subsidies and the Brooke Amendment, 42 U.S.C. sec. 1402(1) which limits rent charges to no more than twenty-five percent of a family's income. In Rhode Island only 3971 public housing units existed in 1972. This figure is less than one-third of the number needed for "AFDC" recipients alone. Furthermore, most of the existing structures are not occupied by them.
The plaintiffs contend that the averaging of these two statistically distinct populations is not "fair averaging" as required by Rosado and reduces and obscures the standard of need for shelter because by averaging the two together the grant for purchase of shelter to recipients living in private housing and paying appreciably more rent than those living in public housing has been artificially lowered below actual need.
I don't question the state's action in this regard. The judiciary's role in determining what equals "fair averaging" and what is the overall effect of the averaging process is not made entirely clear in Rosado v. Wyman, 397 U.S. 397, 419, 90 S. Ct. 1207 (1970). In New Jersey Welfare Rights Organization v. Cahill, supra, the Court rejected a proposal suggesting a qualitative review of the components averaged into a particular consolidated payment. That is to say, relating it to the case at bar, it would not be for this court to determine whether it was proper for the state to average together public and private housing. Rather, as held in Cahill, supra, it would be left to consider only the "technical statistical process" by which the flat grant was computed. In other words, the court in its review would look to see only whether any items have been omitted or whether any need item has been given excess weight in the averaging process. Therefore, the defendant's inclusion of public housing would not be subject to review by this court.
I agree that it is improper for the judiciary to qualitatively evaluate each component of the average. However, one further step must be considered. The federal court cannot totally abdicate its responsibility to insure that the "standard of need" has not been obscured, lowered, or averaged so as to create substantial inequality. As the Court wrote in Rosado, supra:
"While we view with concern the escalating involvement of federal courts in *44 this highly complicated area of welfare benefits, one that should be formally placed under the supervision of HEW, at least in the first instance, we find not the slightest indication that Congress meant to deprive federal courts of their traditional jurisdiction to hear and decide federal questions in this field. It is, of course, no part of the business of this Court to evaluate, apart from federal constitutional or statutory challenge, the merits or wisdom of any welfare programs, whether state or federal, in the large or in the particular. It is, on the other hand, peculiarly part of the duty of this tribunal, no less in the welfare field than in other areas of the law, to resolve disputes as to whether federal funds allocated to the States are being expended in consonance with the conditions that Congress has attached to their use. As Mr. Justice Cardozo stated, speaking for the Court in Helvering v. Davis, 301 U.S. 619, 645 [57 S. Ct. 904, 910, 81 L. Ed. 1307] (1937): `When [federal] money is spent to promote the general welfare, the concept of welfare or the opposite is shaped by Congress, not the states.' Cf. Lassen v. Arizona ex rel. Arizona Highway Dept., 385 U.S. 458 [87 S. Ct. 584, 17 L. Ed. 2d 515] (1967)." 397 U.S. at 422-423, 90 S. Ct. at 1222.
Unlike Cahill this court feels that additionally it must look at the end product of the averaging process to determine whether the figures produced by the averaging are consistently and materially out of proportion with the relevant former definition of "standard of need". See, Rosado v. Wyman, 322 F. Supp. 1173, 1190 (E.D.N.Y.1970) aff'd 437 F.2d 619 (2d Cir.), aff'd 402 U.S. 991, 91 S. Ct. 2169, 26 L. Ed. 2d 157 (1971). This is not to say that a Court can or should look to each component averaged to see if it represents the typical "standard of need". The additional step I suggest provides only an outer check to assure that the averaging process or the inclusion of a certain component in the process does not create a broad distortion in the "standard of need". This limited review beyond that allowed in the Cahill case is essential to discharge the Rosado mandate and HEW regulation, 45 C.F.R. sec. 233.20(a) (1), which provides that "the determination of need and amount of assistance for all applicants and recipients will be made on an objective and equitable basis. . . ." That being the case, further discussion is necessary.
The averaging of statistically distinct populations, such as those living in public housing versus those living in private housing, does not automatically distort the "standard of need" in violation of the fair averaging requirement. I come to this conclusion because averaging of statistically distinct populations was implicitly approved in Rosado v. Wyman, 397 U.S. 397, 90 S. Ct. 1207 (1970) by the Court's acceptance of the averaging of special needs across the entire "AFDC" population despite the fact that many previously received no allocation for special needs. Subsequent to the U. S. Supreme Court opinion in Rosado, the lower court on remand expressly approved the averaging of statistically distinct populations. Rosado v. Wyman, 322 F. Supp. 1173, supra. We also find that in Johnson v. White, 353 F. Supp. 69 (D.Conn.1972), a similar conclusion was reached. Both cases recognized that a merger disadvantages one segment and advantages another. A raise for some and decrease for others is the natural product of any averaging system. If the averaging of populations with different characteristics were objectionable per se, no averaging could take place as distinctions can always be *45 found. On the other hand, as this court has stated, the averaging or merging of separate categories of people is not acceptable where a substantial distortion of the typical "standard of need" is created. HEW clearly does not prohibit the averaging of subsidized public housing units with private housing units, but neither does HEW preclude a state from excluding subsidized housing from its average. Now, it is true that the holding in Johnson v. White, supra, implicitly approving the averaging of the same components, is not directly on point. In Johnson, the "standard of need" prior to the institution of the flat grant system was already related to the cost of public housing and the quantum difference in rental costs between the two was substantially less than appears to be the case here. However, though this distinction exists, it does not change the end result. As will be discussed, infra, the distortion in this case as was the situation in Johnson is not of such proportions as to preclude the inclusion of public housing in the averaging process.
Neither party introduced evidence definitively establishing the number of "AFDC" recipients living in public housing or the rental price differential between public and private housing. However, both sides agreed that roughly 1,000 out of 14,000 assistance groups reside in public housing and that the approximate variation in rent is between fifty to sixty dollars. Using those approximations the averaging of public housing with private housing results in somewhat less than five dollars per assistance unit per month decrease for those living in non-public housing. Thus by itself the averaging together of public and private recipient groups in Rhode Island does not create a broad distortion in the resulting flat grant. To a welfare recipient four to five dollars a month is certainly not de minimus, but in light of the court's role as discussed, supra, accepting the averaging of statistically distinct populations, the quantum of distortion found here is insufficient to bar the averaging of the components in question. On the other hand, had the number of "AFDC" recipients living in public housing equalled or exceeded those in private housing, the result most probably would be different, for in such situation there would, indeed, be a broad distortion. The fact remains that the reality of statistical analysis and the discretion given to the states by Rosado leads to no other conclusion. In short, the court's role in fair averaging is limited as distinguished from fair pricing discussed, infra. However, I add I so find for the purposes of this preliminary hearing. When the case is reached on the merits it may well be that further evidentiary developments by the plaintiffs as to factual realities affecting the recipient may prove a five dollar differential to indeed be a broad distortion.
The plaintiffs further argue that the failure to satisfy public and private housing in computing the flat grant violates 42 U.S.C. sec. 602(a) (1), 45 C.F.R. sec. 233.20(a) (i) and 45 C.F. R. sec. 233.20(a) (2) (b) (iii) which require that a state's "standard of need" be made on an objective, equitable basis and be uniformly applied throughout the state. Plaintiffs contend that the uniformity requirement of the statute means uniformity of purchasing power for each assistance unit. In taking this position they rely heavily on Boddie v. Wyman, 434 F.2d 1207 (2d Cir. 1970) aff'd 402 U.S. 991, 91 S. Ct. 2168, 29 L. Ed. 2d 157 (1971). I don't share such confidence in the Boddie holding. True, it does hold that assistance payment differentials to "AFDC" recipients are not acceptable unless they can be factually justified. However, it dealt with a situation involving geographic differentials and its discussion indicates that the provisions at issue were intended specifically to assure intrastate geographic uniformity. *46 Id. 434 F.2d at 1210, 1211. The issue of geographic uniformity is not presented here, a point not considered by the plaintiffs. If plaintiffs' position is carried to the extreme, this argument runs directly in the face of the concept of averaging accepted in Rosado v. Wyman, 397 U.S. 397, 90 S. Ct. 1207 (1970). The need for uniform, objective, equitable assistance payments must be read in relation to the holding in Rosado permitting averaging. Therefore, although a certain amount of inequality is unavoidable in any averaging process, where the criteria enunciated in Rosado are met, no violation exists.
Next the plaintiffs maintain that the combined averaging of public and private housing invidiously discriminates against the plaintiff class with no rational basis, in violation of the Fourteenth Amendment to the United States Constitution because the flat grant provides more shelter related income to those in public housing than is needed, while providing less than is needed for those in private housing.
This raises the constitutional issue which I need not reach and which the parties have agreed be deferred.
Averaging of Pro-Rated Units
Plaintiffs claim that defendant's policy at the time American Data Systems, Inc. computed the flat grant, was to reduce the payment for shelter in "non-square" "AFDC" cases based upon an irrebuttable presumption that the nonrecipient or "MARS" contributed to the household expenses an amount equal to his or her pro-rata share of shelter costs. Plaintiffs do not contend that in any instance the averaging of "non-square" units is improper. Rather plaintiffs assert that the irrebuttable presumption violates federal law and consequently cannot be used in the averaging process. This Court agrees. The use of grants based upon irrebuttable presumptions in the computation of the consolidated standard is impermissible.
A policy which conclusively presumes the availability of income to a recipient by one not legally obligated to contribute such support is in direct conflict with 45 C.F.R. sec. 233.20 (a) (3) (ii) (c), which provides:
". . . in establishing financial eligibility and the amount of assistance payment: only such net income as is actually available for current use on a regular basis will be considered, and only currently available resources will be considered;" (emphasis added)
Moreover, while approving reasonable, rebuttable presumptions, courts have uniformly struck down regulations based upon unverified, conclusive presumptions of income. See, Lewis v. Martin, 397 U.S. 552, 90 S. Ct. 1282, 25 L. Ed. 2d 561 (1970); Rodriguez v. Vowell, 472 F.2d 622, 627 (5th Cir. 1973); Gilliard v. Craig, 331 F. Supp. 587 (W.D.N.C. 1971), aff'd 409 U.S. 807, 93 S. Ct. 39, 34 L. Ed. 2d 66 (1972); Solman v. Shapiro, 300 F. Supp. 409 (D.Conn.), aff'd 396 U.S. 5, 90 S. Ct. 25, 24 L. Ed. 2d 5 (1969); Owens v. Parham, 350 F. Supp. 598 (N. D.Ga., Atlantic Div.1972).
In factual support of their position the plaintiffs point to sec. 11, E, 3 of defendant's policy manual in effect at time of the averaging, which on its face indicates that a stepfather or "MARS" is always responsible for the entire support of the woman with whom he is living. In addition, plaintiffs presented Robert Parry, Social Service Supervisor, Department of Social and Rehabilitative Services as a witness, who, although not *47 intimately involved with the day to day workings of the assistance payment or the Fair Hearing process, testified that the presumption of income support for the woman when living with a "MARS" was viewed by the department as an irrebuttable presumption. The defense contends this is not so because the agency provides an administrative hearing process to any recipient dissatisfied with any agency decision. This is of no account, for there is no evidence that this hearing can be used to rebut a presumption of income if in fact the agency policy is to the contrary. On the premise that the agency policy does create an irrebuttable presumption then, of course, the hearing mechanism is not for the purpose of changing such policy.
While the quantum of evidence presented is not sufficient for a final determination that defendant's policy creates an irrebuttable presumption, I feel that the requisite burden for this stage of the proceedings, to demonstrate a strong likelihood of success on the merits has been met. Furthermore, while the plaintiffs were unable to present evidence to indicate the exact number of individuals who at the time of the survey were affected by the policy, it is clear that the policy had a depressing effect on the grant level of those recipients affected and would thus obscure and lower the standard of need impermissibly in the consolidated grant.
Averaging of Rental Costs in November 1972
The plaintiffs challenge the averaging of rental costs during a period of economic controls with no provision for updating the figures at the time of implementation, to take into account increases due to inflation. The survey conducted by American Data Systems, Inc. was conducted in November 1972 whereas the flat grant was not instituted until November 1973. At the time the averaging took place, rental charges were being strictly regulated, although not frozen, pursuant to 6 C.F.R. sec. 301 et seq. In the interim rental charges have increased markedly.
As I have previously pointed out Rosado requires that in converting or as is said "redefin[ing] its method for determining need" a state must provide that ". . . all factors in the old equation [be] accounted for and fairly priced and [provide that] the consolidation [be] on a statistical basis reflect[ing] a fair averaging." The thrust of the problem is whether or not the November 1972 rental cost figures used by American Data Systems fairly prices the rent factor. The answer is no. It follows that where a state purports to have a "standard of need" based on actual need, a consolidation based upon costs a year old violates the Rosado requirement and obscures the state's articulated "standard of need".
According to the Court in New Jersey Welfare Rights Organization v. Cahill, supra, the term "fairly priced" means "what the price actually is" and does not mean "what the price would be by market standards or what would be required for a decent existence". 349 F. Supp. at 511. Refining its definition the court wrote:
"By the words `fairly priced' the Supreme Court was apparently trying to prevent a state, barred from totally eliminating a factor from the content of its former standard of need, from artificially pricing an item so low that *48 that item would essentially be eliminated in terms of reality. For example, a special circumstance item such as a telephone, normally priced at about $4.50 per month, cannot be represented in the averaging process at a cost of $1.00 per year. It would not be a fully considered need item. See, Rhode Island Fair Welfare Rights Organization v. Department of Social and Rehabilitative Services, 329 F. Supp. 860, 867 (D.R.I.1971)." 349 F. Supp. at 511.
Any other definition would fundamentally subvert the integrity of 42 U.S.C. sec. 602 (a) (23) as interpreted by Rosado. A "Standard of Need" articulated as actual need prior to the conversion to a consolidated grant must, therefore, be based upon up-to-date prices as of the date of implementation, or else the "standard of need" as of the date of implementation cannot reflect actual need. New Jersey Welfare Rights Organization v. Cahill, supra, 349 F. Supp. at 512. In conclusion, this Court finds that the State of Rhode Island must take into account inflation from November 1972 to November 1973 in its redefined "standard of need".
Here again the criteria for probability of success on the merits has been satisfied.
Failure to Survey and/or Average Actual Rents
The plaintiffs' next major claim relates to the review and averaging of budgeted rents only in computing the redefined standard of need. Defendant concedes that no survey was taken of actual rental costs. Plaintiffs argue that although the state's policy was by and large to pay "actual need", in fact, this policy was not meticulously carried out and budgeted rental allotments often were less than actual need. As a consequence, the averaging of budgeted rents does not equal "fair pricing" and the resultant average obscures and lowers the "standard of need".
As has previously been discussed, "fair pricing" means "cost as is" and when a state previously purported to meet actual need, cost equals need. Therefore, if an item is accounted for in the averaging at less than actual cost, the "standard of need" would be impermissibly "lowered and obscured". New Jersey Welfare Rights Organization v. Cahill, supra. This Court has already found that the defendant's policy was, for all practial purposes, to pay rental cost as actually needed. However, defendant's own expert witness, Mr. Richard Gertzof of American Data Systems, Inc., expressed surprise when he discovered that a great many recipients were budgeted less than actual need. The local welfare director for the City of Cranston testified that well over a third of the city's caseload received less than actual rental cost. No limitation based upon reasonableness could account for this discrepancy.
In any case this Court believes that a survey of actual rents is necessary to verify the accuracy of budgeted rents. For purposes of determining the "standard of need" and determining to what extent budgeted rents fell short of actual rents, a separate housing survey was taken in Johnson v. White, supra. Only upon ascertaining that the budgeted rents accurately reflected actual rents did the Court in Johnson approve the use of budgeted rents as fair pricing in computation of the redefined "standard of need". No such survey was conducted in Rhode Island and the evidence indicates that budgeted rents here may vary markedly from actual rents. Therefore, this Court holds that the averaging of unverified budgeted rents here involved a violation of the "fair pricing" requirement of Rosado creating a situation whereby the announced policy at the time of conversion was to pay rent as needed subject only to a reasonableness limitation, but the reality of the figures used appear to have fallen significantly below that norm. In this *49 manner the extent to which the state met its announced "standard of need" has been obscured.
Having found that the plaintiff class has a probability of success on the merits on several claims, analysis of the equitable interests of both parties for purposes of determining the appropriateness of a preliminary injunction is necessary. In September 1973, 13,268 family groups in Rhode Island participated in the "AFDC" program. Of this number, John Affleck, Director of the Department of Social and Rehabilitative Services, testified that some 35% would receive decreases, and 65% would receive increases as a result of the flat grant. However, on cross-examination these figures were shown to be inaccurate and the percentage of families receiving decreases appears to be larger. Among those receiving decreases, almost all were previously budgeted more for purchase of shelter than they are receiving under the flat grant. Directors of three local assistance offices corroborated affidavits of several "AFDC" recipients that housing at the cost budgeted in the flat grant is extraordinarily difficult to locate.
For these individuals the decrease in "AFDC" payments comes at a time of sharply spiralling costs in all sectors of life. Food and fuel costs have risen substantially in 1973. While plaintiffs failed to present evidence concerning the number of welfare recipients threatened with the loss of their present shelter or the cut-off of heating fuel, sufficient evidence was presented to indicate that the crisis for many this winter will be grave indeed. The crisis to those receiving a decrease is further enhanced because under the flat grant system, those unable to meet their utility bills will be unable to turn to the State for special help as they previously could. The fact that special needs have been averaged into the overall flat grant offers little solace to those receiving a decrease in their assistance payment.
Defendant counters that any money expended by the State under an injunction will be irretrievably lost. Second, defendant notes that if the State decides not to spend additional revenues, the injunction will cause a decrease in benefits to those "AFDC" recipients now living in public housing and not part of the plaintiff class. This Court concludes that the compromise suggested by the plaintiffs offers the greatest opportunity to avoid undue harm to all concerned. Therefore, it is hereby ordered that the defendant be and is hereby enjoined from using as its means of assistance payment to "AFDC" recipients the averages comprising that portion of the *50 flat grant allocated to the purchase of shelter.
It is further ordered that in place thereof, purchase of shelter will be paid on an "as need" basis in accordance with the policy existing prior to November 1, 1973.
This matter came to be heard on February 19, 1974 concerning plaintiffs' motion pursuant to Rule 52(b) and Rule 60(a) of the Federal Rules of Civil Procedure to add to the findings of fact contained in the court's opinion and order of February 11, 1974. After consideration of the motion and memorandum in support thereof, the argument of counsel; a careful review of the testimony and evidence involved and finding said testimony and evidence sufficient, it is hereby
Ordered, Adjudged and Decreed that the Opinion and Order of February 11, 1974 shall be amended insofar as the following additional findings of fact are made:
Based upon the testimony of plaintiffs' witnesses, particularly the testimony of the Directors of the Welfare Department of the Cities of East Greenwich, East Providence, and Cranston, and upon the affidavits of members of plaintiffs class which were admitted into evidence with the consent of the defendant, I find a substantial probability that, absent the relief granted herein, many of the plaintiffs and members of plaintiff's class may be forced, while this action is pending, to abandon their rented or mortgage-self-owned homes and even to foresake certain communities because of the substantial reductions in their housing allowances under the averaging plan implemented by the defendant and because of the absence of available low-priced rental housing in many communities in the State.NOTES
 A "non-square" or pro-rated family is one in which not all members in the household are "AFDC" recipients. A "square" family is one in which the number of "AFDC" recipients is equal to the number of persons in the household.
 The "special needs" columns were formerly paid as special needs and not as part of the basic assistance payment for rent. They reflect an averaging of the amount the State used to pay to help a recipient avert eviction in instances where need exceeded payment and a recipient was unable to meet his obligations. Under the new system these special needs are averaged into the flat grant and will not be met on an individual as need basis. Mr. Soucy of American Data Systems, Inc. testified that while 90% of these figures is attributable to heat, rent, or utilities, how much specifically went to back rent is uncertain.
 The stepfather, "MARS", column accounts for an adjustment made subsequent to the completion of the original flat grant survey to take into consideration the overhead needs of children when there was a stepfather or Man Assuming the Role of Spouse ("MARS") in the household. A small portion of this amount is traceable to heat and utilities rather than rent.
 Rhode Island's policy, concerning shelter or rent costs as of the date of the survey for purposes of computing the flat grant, was taken as set out in the Rhode Island Department of Social and Rehabilitative Services, Manual of Assistance Payment for Families at Section 11, page 5 as follows:
Shelter is purchased in various ways, with costs met according to the way this requirement is purchased.
Rent, as paid may be approved by the case aide up to $80.00 a month. Rent between $80.00 and $120.00 a month is referred to the Casework Supervisor for evaluation and decision. A AP-81 is required. Rent above $120.00 a month requires the approval of the Senior Supervisor.
Approval for rent in excess of $80.00 a month is related to the size and needs of the family, the number of rooms required, the prevailing rates in the community and local minimum housing standards."
In the Technical Report prepared by American Data Systems, Inc., the conclusion was expressly reached that the figures, contained in the above policy, acted solely as administrative levels and not shelter maximums. (Plaintiffs' exhibit #9 P. 24-25)
 This finding comports with that of the Technical Report prepared by American Data Systems, Inc. at page 17 which deserves great weight although A.D.S. representative Roger Soucy's subsequent testimony on the same point indicates a somewhat inconsistent conclusion. Mr. Joseph Murray, the Administrator of the Assistance Payments Division of the Department of Social and Rehabilitative Services testified that the standard of need in Rhode Island is traditionally actual cost depending on the needs of the client. From the whole of his testimony I conclude that while actual cost was not paid in each instance, by and large actual cost was the standard.
Moreover, this Court on a previous occasion found that actual expenses for shelter costs were budgeted for "AFDC" recipients, Rhode Island Fair Welfare Rights Org. v. Dept. of Social and Rehabilitative Services, 329 F. Supp. 860, 863 (D.R.I.1971).
 Directory "Low Rent Public Housing in Rhode Island", Rhode Island Department of Community Affairs, December 1972.
 Prior to the institution of the consolidated assistance payment in Connecticut, the Standard of Need for recipients living in private housing was budgeted on the basis of the cost of a comparable unit of public housing, plus 10% up to a dollar maximum, with an additional 10% allotted for furnished housing. Johnson v. White, 353 F. Supp. at 77.
 45 C.F.R. 233.20(a) (2) (b) (iii): "Provides that the standard will be uniformly applied throughout the state;"
 This section reads:
"3. Shelter Costs of a Child living with a Stepfather or with an Unrelated Man assuming Role of Spouse.
When a child lives with a mother and stepfather or, with an unrelated man assuming the role of spouse to the mother, the financial responbibility of the stepfather or the unrelated man is as follows:
* * * * *
c. The man is always responsible for the entire support of his wife or for the woman with whom he is living and assuming the role of spouse."
 According to the Boston Consumer Price Indexes for July 1973 and October 1972, housing costs on a 1967 base of 100 are as follows:
Index Index July 1973 Oct. 1972 Shelter 145.8 143.5 Rent 136.0 131.2 Home ownership 150.0 149.0
 By this the court meant that the state need not eliminate an element such as public housing which is artificially priced below the market standard as a result of government intervention.
 The figures upon which Mr. Affleck and subsequently Ruth Coogan, Chief of Division of Research and Statistics for the Department of Social and Rehabilitative Services, based their estimate of the percentage of families benefited and harmed by the conversion did not take into account non-recurring special need payments under the old system, although this component was added into the flat grant. If this component had been taken into account, the total grant under the old system would have been greater. The failure to properly compare like components resulted in a distorted comparison of the effect of the two systems.
I also conclude that the testimony of Sister Jeanine Maynard, a statistical expert, called by the plaintiffs appear to contain errors. She stated that closer to 47% received decreases. The figure should be somewhere in between.
 Sister Jeanine Maynard testified that in the three communities for which she had information, 642 of 644 families receiving a decrease under the flat grant were previously budgeted more for shelter than is built into the flat grant.
 In determining the hardship caused by the flat grant the court has not taken into account the figures cited by the plaintiff from the "1970 Census of Housing for Rhode Island" published by the United States Department of Commerce, Bureau of the Census because the figures cited relate to median gross rent. Gross rent includes fuel and utility costs as well as purchase of rent costs.