JEFFERSON v. HACKNEY

MR. JUSTICE REHNQUIST delivered the opinion of the Court.

Appellants in this case challenge certain computation procedures that the State of Texas uses in its federally assisted welfare program. Believing that neither the Constitution nor the federal welfare statute prohibits the State from adopting these policies, we affirm the judgment of the three-judge court below upholding the state procedures.

I

Appellants are Texas recipients of Aid to Families With Dependent Children (AFDC). They brought two class actions, which were consolidated in the United States District Court for the Northern District of Texas, seeking injunctive and declaratory relief against state welfare officials. A three-judge court was convened pursuant to 28 U. S. C. § 2281.

The Texas State Constitution provides a ceiling on the amount the State can spend on welfare assistance grants.1 In order to allocate this fixed pool of welfare money among the numerous individuals with acknowledged need, the State has adopted a system of percentage grants. Under this system, the State first computes the monetary needs of individuals eligible for relief under each of the federally aided categorical assistance programs.2 Then, since the constitutional ceiling on welfare is insufficient to bring each recipient up to this full standard of need, the State applies a percentage reduction factor3 in order to arrive at a reduced standard of need in each category that the State can guarantee.

Appellants challenge the constitutionality of applying a lower percentage reduction factor to AFDC than to the other categorical assistance programs. They claim a violation of equal protection because the proportion of AFDC recipients who are black or Mexican-American is higher than the proportion of the aged, blind, or disabled welfare recipients who fall within these minority groups. Appellants claim that the distinction between the programs is not rationally related to the purposes of the Social Security Act, and violates the Fourteenth Amendment for that reason as well. In their original complaint, appellants also argued that any percentage-reduction system violated § 402 (a) (23) of the Social Security Act of 1935, as amended, 81 Stat. 898, 42 U. S. C. § 602 (a) (23), which required each State to make certain cost-of-living adjustments to its standard of need.

The three-judge court rejected appellants' constitutional arguments, finding that the Texas system is neither racially discriminatory nor unconstitutionally arbitrary. The court did, however, accept the statutory claim that Texas' percentage reductions in the AFDC program violate the congressional command of § 402 (a) (23). 304 F.Supp. 1332 (ND Tex. 1969).

Subsequent to that judgment, this Court decided Rosado v. Wyman, 397 U.S. 397 (1970). Rosado held that, although § 402 (a) (23) required States to make cost-of-living adjustments in their standard-of-need calculations, it did not prohibit use of percentage-reduction systems that limited the amount of welfare assistance actually paid. 397 U. S., at 413. This Court then vacated and remanded the first Jefferson judgment for further proceedings consistent with Rosado. 397 U.S. 821 (1970).

On remand, the District Court entered a new judgment, denying all relief. Then, in a motion to amend the judgment, appellants raised a new statutory claim. They argued for the first time that although a percentage-reduction system may be consistent with the statute, the specific procedures that Texas uses for computing that reduction violate the congressional enactment. The District Court rejected this argument and denied without opinion appellants' motion to amend the judgment. This appeal under 28 U. S. C. § 1253 then followed, and we noted probable jurisdiction. 404 U.S. 820 (1971).

II

Appellants' statutory argument relates to the method that the State uses to compute the percentage reduction when the recipient also has some outside income. Texas, like many other States,4 first applies the percentage-reduction factor to the recipient's standard of need, thus arriving at a reduced standard of need that the State can guarantee for each recipient within the present budgetary restraints. After computing this reduced standard of need, the State then subtracts any nonexempt5 income in order to arrive at the level of benefits that the recipient needs in order to reach his reduced standard of need. This is the amount of welfare the recipient is given.

Under an alternative system used by other States, the order of computation is reversed. First, the outside income is subtracted from the standard of need, in order to determine the recipient's "unmet need." Then, the percentage-reduction factor is applied to the unmet need, in order to determine the welfare benefits payable.

The two systems of accounting for outside income yield different results.6 Under the Texas system all welfare recipients with the same needs have the same amount of money available each month, whether or not they have outside income. Since the outside income is applied dollar for dollar to the reduced standard of need, which the welfare department would otherwise pay in full, it does not result in a net improvement in the financial position of the recipient. Under the alternative system, on the other hand, any welfare recipient who also has outside income is in a better financial position because of it. The reason is that the percentage-reduction factor there is applied to the "unmet need," after the income has been subtracted. Thus, in effect, the income-earning recipient is able to "keep" all his income, while he receives only a percentage of the remainder of his standard of need.7

Each of the two systems has certain advantages. Appellants note that under the alternative system there is a financial incentive for welfare recipients to obtain outside income. The Texas computation method eliminates any such financial incentive, so long as the outside income remains less than the recipient's reduced standard of need.8 However, since Texas' pool of available welfare funds is fixed, any increase in benefits paid to the working poor would have to be offset by reductions elsewhere. Thus, if Texas were to switch to the alternative system of recognizing outside income, it would be forced to lower its percentage-reduction factor, in order to keep down its welfare budget. Lowering the percentage would result in less money for those who need the welfare benefits the most—those with no outside income—and the State has been unwilling to do this.

Striking the proper balance between these competing policy considerations is, of course, not the function of this Court. "There is no question that States have considerable latitude in allocating their AFDC resources, since each State is free to set its own standard of need and to determine the level of benefits by the amount of funds it devotes to the program." King v. Smith, 392 U.S. 309, 318-319 (1968) (footnotes omitted).9 So long as the State's actions are not in violation of any specific provision of the Constitution or the Social Security Act, appellants' policy arguments must be addressed to a different forum.

Appellants assert, however, that the Texas computation procedures are contrary to § 402 (a) (23):

.....

Recognizing that this statutory language, by its terms, hardly provides much support for their theory, appellants seek to rely on what they perceive to have been the broad congressional purpose in enacting the provision.

In Rosado v. Wyman, supra, the Court reviewed the history of this section and rejected the argument that it had worked any radical shift in the AFDC program. Id., at 414 and n. 17. AFDC has long been referred to as a "scheme of cooperative federalism," King v. Smith, 392 U. S., at 316, and the Rosado Court dismissed as "adventuresome" any interpretation of § 402 (a) (23) that would deprive the States of their traditional discretion to set the levels of payments. 397 U. S., at 414-415 and n. 17. Instead, the statute was meant to require the States to make cost-of-living adjustments to their standards of need, thereby serving "two broad purposes":

Texas has complied with these two requirements. Effective May 1, 1969, the standard of need for AFDC recipients was raised 11% to reflect the rise in the cost of living, and the State shifted from a maximum-grant system to its present percentage-reduction system. In this way, the State has fairly recognized and exposed the precise level of unmet need, and by using a percentage-reduction system it has attempted to apportion the State's limited benefits more equitably.

Although Texas has thus responded to the "two broad purposes" of § 402 (a) (23), appellants argue that Congress also intended that statute to increase the total number of recipients of AFDC, so that more people would qualify for the subsidiary benefits that are dependent on receipt of AFDC cash assistance.10 The Texas computation procedures are thought objectionable since they do not increase the welfare rolls to quite the same extent as would the alternative method of recognizing outside income.

We do not agree that Congress intended § 402 (a) (23) to invalidate any state computation procedures that do not absolutely maximize individual eligibility for subsidiary benefits. The cost-of-living increase that Congress mandated would, of course, generally tend to increase eligibility,11 but there is nothing in the legislative history indicating that this was part of the statutory purpose. Indeed, at the same time Congress enacted § 402 (a) (23) it included another section designed to induce States to reduce the number of individuals eligible for the AFDC program.12 Thus, what little legislative history there is on the point, see Rosado v. Wyman, 397 U. S., at 409-412, tends to undercut appellants' theory. See Lampton v. Bonin, 304 F.Supp. 1384, 1391-1392 (ED La. 1969) (Cassibry, J., dissenting). See generally Note, 58 Geo. L. J. 591 (1970).

Appellants also argue that the Texas system should be held invalid because the alternative computation method results in greater work incentives for welfare recipients.13 The history and purpose of the Social Security Act do indicate Congress' desire to help those on welfare become self-sustaining. Indeed, Congress has specifically mandated certain work incentives in § 402 (a) (8). There is no dispute here, however, about Texas' compliance with these very detailed provisions for work incentives. Neither their inclusion in the Act nor the language used by Congress in other sections of the Act supports the inference that Congress mandated the States to change their income-computation procedures in other, completely unmentioned areas.

Nor are appellants aided by their reference to Social Security Act § 402 (a) (10), 42 U. S. C. § 602 (a) (10), which provides that AFDC benefits must "be furnished with reasonable promptness to all eligible individuals." That section was enacted at a time when persons whom the State had determined to be eligible for the payment of benefits were placed on waiting lists, because of the shortage of state funds. The statute was intended to prevent the States from denying benefits, even temporarily, to a person who has been found fully qualified for aid. See H. R. Rep. No. 1300, 81st Cong., 1st Sess., 48, 148 (1949); 95 Cong. Rec. 13934 (remarks of Rep. Forand). Section 402 (a) (10) also prohibits a State from creating certain exceptions to standards specifically enunciated in the federal Act. See, e. g., Townsend v. Swank, 404 U.S. 282 (1971). It does not, however, enact by implication a generalized federal criterion to which States must adhere in their computation of standards of need, income, and benefits.14 Such an interpretation would be an intrusion into an area in which Congress has given the States broad discretion, and we cannot accept appellants' invitation to change this longstanding statutory scheme simply for policy consideration reasons of which we are not the arbiter.

III

We turn, then, to appellants' claim that the Texas system of percentage reductions violates the Fourteenth Amendment. Appellants believe that once the State has computed a standard of need for each recipient, it is arbitrary and discriminatory to provide only 75% of that standard to AFDC recipients, while paying 100% of recognized need to the aged, and 95% to the disabled and the blind. They argue that if the State adopts a percentage-reduction system, it must apply the same percentage to each of its welfare programs.

This claim was properly rejected by the court below. It is clear from the statutory framework that, although the four categories of public assistance found in the Social Security Act have certain common elements, the States were intended by Congress to keep their AFDC plans separate from plans under the other titles of the Act.15 A State is free to participate in one, several, or all of the categorical assistance programs, as it chooses. It is true that each of the programs is intended to assist the needy, but it does not follow that there is only one constitutionally permissible way for the State to approach this important goal.

This Court emphasized only recently, in Dandridge v. Williams, 397 U.S. 471, 485 (1970), that in "the area of economics and social welfare, a State does not violate the Equal Protection Clause merely because the classifications made by its laws are imperfect." A legislature may address a problem "one step at a time," or even "select one phase of one field and apply a remedy there, neglecting the others." Williamson v. Lee Optical Co., 348 U.S. 483, 489 (1955). So long as its judgments are rational, and not invidious, the legislature's efforts to tackle the problems of the poor and the needy are not subject to a constitutional straitjacket. The very complexity of the problems suggests that there will be more than one constitutionally permissible method of solving them.

The standard of judicial review is not altered because of appellants' unproved allegations of racial discrimination. The three-judge court found that the "payment by Texas of a lesser percentage of unmet needs to the recipients of the AFDC than to the recipients of other welfare programs is not the result of racial or ethnic prejudice and is not violative of the federal Civil Rights Act or the Equal Protection Clause of the 14th Amendment." The District Court obviously gave careful consideration to this issue, and we are cited by its opinion to a number of subsidiary facts to support its principal finding quoted above. There has never been a reduction in the amount of money appropriated by the legislature to the AFDC program, and between 1943 and the date of the opinion below there had been five increases in the amount of money appropriated by the legislature for the program, two of them having occurred since 1959.16 The overall percentage increase in appropriation for the programs between 1943 and the time of the District Court's hearing in this case was 410% for AFDC, as opposed to 211% for OAA and 200% for AB. The court further concluded:

Appellants in their brief in effect abandon any effort to show that these findings of fact were clearly erroneous, and we hold they were not.

Appellants are thus left with their naked statistical argument: that there is a larger percentage of Negroes and Mexican-Americans in AFDC than in the other programs,17 and that the AFDC is funded at 75% whereas the other programs are funded at 95% and 100% of recognized need. As the statistics cited in the footnote demonstrate, the number of minority members in all categories is substantial. The basic outlines of eligibility for the various categorical grants are established by Congress, not by the States; given the heterogeneity of the Nation's population, it would be only an infrequent coincidence that the racial composition of each grant class was identical to that of the others. The acceptance of appellants' constitutional theory would render suspect each difference in treatment among the grant classes, however lacking in racial motivation and however otherwise rational the treatment might be. Few legislative efforts to deal with the difficult problems posed by current welfare programs could survive such scrutiny, and we do not find it required by the Fourteenth Amendment.18

Applying the traditional standard of review under that amendment, we cannot say that Texas' decision to provide somewhat lower welfare benefits for AFDC recipients is invidious or irrational. Since budgetary constraints do not allow the payment of the full standard of need for all welfare recipients, the State may have concluded that the aged and infirm are the least able of the categorical grant recipients to bear the hardships of an inadequate standard of living. While different policy judgments are of course possible, it is not irrational for the State to believe that the young are more adaptable than the sick and elderly, especially because the latter have less hope of improving their situation in the years remaining to them. Whether or not one agrees with this state determination, there is nothing in the Constitution that forbids it.19

Similarly, we cannot accept the argument in MR. JUSTICE MARSHALL's dissent that the Social Security Act itself requires equal percentages for each categorical assistance program. The dissent concedes that a State might simply refuse to participate in the AFDC program, while continuing to receive federal money for the other categorical programs. See post, at 577. Nevertheless, it is argued that Congress intended to prohibit any middle ground—once the State does participate in a program it must do so on the same basis as it participates in every other program. Such an all-or-nothing policy judgment may well be defensible, and the dissenters may be correct that nothing in the statute expressly rejects it. But neither does anything in the statute approve or require it.20

In conclusion, we re-emphasize what the Court said in Dandridge v. Williams, 397 U. S., at 487:

Affirmed.

MR. JUSTICE STEWART joins in Part III of the Court's opinion.

MR. JUSTICE DOUGLAS, with whom MR. JUSTICE BRENNAN concurs, dissenting.

I would read the Act more generously than does the Court. It is stipulated that 87% of those receiving AFDC aid are blacks or Chicanos. I would therefore read the Act against the background of rank discrimination against the blacks and the Chicanos and in light of the fact that Chicanos in Texas fare even more poorly than the blacks. See L. Grebler, J. Moore, & R. Guzman, The Mexican-American People, pts. 2 and 3 (1970); J. Burma, Mexican-Americans in the United States 143-199 (1970); Schwartz, State Discrimination Against Mexican Aliens, 38 Geo. Wash. L. Rev. 1091 (1970); U. S. Commission on Civil Rights, The Mexican American (1968); U. S. Commission on Civil Rights, Mexican Americans and the Administration of Justice in the Southwest (1970). In Rosado v. Wyman, 397 U.S. 397, 413, we said that in administering such a program a State "may not obscure the actual standard of need." Texas does precisely that by manipulating a mathematical formula.

In Rosado, we described how some States establish upper limits or maximums of aid, while others, like Texas, "curtail the payments of benefits by a system of `ratable reductions' whereby all recipients will receive a fixed percentage of the standard of need." Id., at 409. Then in footnote 13 we described what that meant: "A `ratable reduction' represents a fixed percentage of the standard of need that will be paid to all recipients. In the event that there is some income that is first deducted, the ratable reduction is applied to the amount by which the individual or family income falls short of need." Id., at 409 n. 13 (emphasis added).

If Texas first deducted outside income and then made its ratable reduction, the welfare recipient would receive a somewhat more generous payment, as the opinion of the Court illustrates in footnote 6 of its opinion. Not only does the Texas system avoid this generous approach, but it also impermissibly constricts the standard of need in conflict with Rosado, Dandridge v. Williams, 397 U.S. 471, and Townsend v. Swank, 404 U.S. 282. Under Texas' method of computation, a family—otherwise eligible for AFDC benefits but with nonexempt income greater than the level of benefits and less than the standard of need—is denied both AFDC cash benefits and other noncash benefits such as medicaid.1 It seems inconceivable that Congress could have intended that noncash benefits be denied those with incomes less than the standard of need solely because that income was earned rather than from categorical assistance. Yet this is precisely the result sanctioned by the Court today because eligibility for these programs is tied to the receipt of cash benefits.2

One of the stated purposes of the AFDC program is "to help such parents or relatives [of needy dependent children] to attain or retain capability for the maximum self-support and personal independence." 42 U. S. C. § 601 (emphasis added). The Senate Finance Committee has stated, "A key element in any program for work and training for assistance recipients is an incentive for people to take employment." S. Rep. No. 744, 90th Cong., 1st Sess., 157 (1967) (emphasis added). The majority acknowledges that "[t]he history and purpose of the Social Security Act . . . indicate Congress' desire to help those on welfare become self-sustaining." Ante, at 544. But it nonetheless ignores the explicit congressional policy in favor of work incentives and upholds a system which provides penalties and disincentives for those who seek employment.3

The California Supreme Court in Villa v. Hall, 6 Cal.3d 227, 490 P.2d 1148, struck down the system this Court approves today, where California used a statutory maximum of payments rather than a ratable reduction. The California Supreme Court quite properly said that what the State was attempting was inconsistent with Rosado. Moreover, it had an additional reason:

Moreover, Townsend v. Swank, 404 U.S. 282, calls for a reversal in the present case. It is conceded that plaintiff Maria T. Davilla and 2,470 other families are denied aid in Texas by reason of its new formula, see 304 F.Supp. 1332, 1343, despite the fact that their income is below the standard of need and that of those receiving AFDC aid only 75% of their needs is met.4

Under § 402 (a) (10) of the Social Security Act (which governs AFDC) "aid to families with dependent children shall be furnished with reasonable promptness to all eligible individuals." 42 U. S. C. § 602 (a) (10). In Townsend children 18 through 20 years of age who attended high school or vocational training were eligible for AFDC benefits but such children in college were not eligible. We held that "a state eligibility standard that excludes persons eligible for assistance under federal AFDC standards violates the Social Security Act and is therefore invalid under the Supremacy Clause."5 404 U. S., at 286.

What Texas does here is to exclude large numbers of AFDC beneficiaries by application of a state eligibility test that is narrower than the one we approved in Rosado. While a State has some discretion in its use of federal funds, it may not manipulate by its own formula groups of "needy" claimants. The decision to participate or not in the federal program is left to the States. Townsend v. Swank, supra, at 290-291. When, as here, federal and state funds are in short supply, the problem is not to lop off some categories of those in "need" but to design a way of managing the system of "need" so as not to raise equal protection questions.6 Id., at 291.

Section 402 (a) (10) of the Social Security Act provides that AFDC shall be furnished with reasonable promptness to all eligible individuals. The House Report in commenting on it said:

As the Court said in Dandridge v. Williams, 397 U. S., at 481, "So long as some aid is provided to all eligible families and all eligible children, the statute itself is not violated." It is violated here because nearly 2,500 families that satisfy the requirements of "need" are denied any relief.7

MR. JUSTICE MARSHALL, with whom MR. JUSTICE BRENNAN joins, and with whom MR. JUSTICE STEWART joins as to Part I only, dissenting.

Appellants, recipients of Aid to Families With Dependent Children (AFDC) in Texas, brought this action to challenge two distinct aspects of the Texas AFDC program. First, appellants challenge the manner in which Texas arrives at the amount it will pay to persons who are needy. Second, they urge that Texas acts illegally in providing more money for persons receiving aid under other social welfare legislation than for persons receiving AFDC aid. The Court rejects both claims. I dissent.

Before proceeding to explain why I disagree with the Court, I would like to illustrate what the disputes in this case are all about. If a State is unable or unwilling to establish a level of AFDC payments to meet all the needs of all recipients, federal law permits the State to use a percentage-reduction factor as a method of reducing payments in a somewhat equitable manner. Texas has adopted a system in which the percentage-reduction factor is applied against the standard of need before outside income is deducted. Appellants contend that federal law requires the State to deduct outside income before the percentage-reduction factor is applied. While describing the differences between the two alternatives is a Herculean task, the figures themselves are not difficult to comprehend. Footnote 6 of the Court's opinion, for example, demonstrates that the Texas system provides less aid to a family with outside income than the alternative system. It is also immediately obvious that under the Texas system, as soon as the family's income reaches $150, it no longer receives anything from the State, whereas under the alternative, a family earning the same $150 would continue to receive some state funds. Hence, the Texas method of computation contracts the class of families eligible to receive state aid. Appellants contend that the characteristics of the Texas system are inconsistent with federal legislation and that only the alternative system comports with the intent of Congress. I agree.

Appellants also claim that the percentage-reduction factor employed by Texas is illegal, irrespective of the method of computing payments, because it is lower than the factor used in other social welfare programs that have participants with identical standards of need. I also agree with appellants on this point, but for slightly different reasons from those they have urged.

I

A. In considering the question whether Texas' method of computing eligibility for AFDC payments comports with the federal statute, 42 U. S. C. § 601 et seq., it is important to keep in mind the words of Mr. Justice Cardozo: "When [federal] money is spent to promote the general welfare, the concept of welfare or the opposite is shaped by Congress, not the states." Helvering v. Davis, 301 U.S. 619, 645 (1937). Mr. Justice Harlan reiterated this point in Rosado v. Wyman, 397 U.S. 397, 422-423 (1970), when he stated that irrespective of the policies that a State might wish to pursue by utilizing AFDC money in one way or another, the ultimate question to be answered in each case is whether the action of the State comports with the requirements of federal law.

The Court concludes in the instant case that there is no general congressional policy violated by Texas' choice between the alternative methods of applying a percentage-reduction factor to its determined standard of need, and also that no specific statutory provision prohibits Texas from choosing one alternative rather than the other. In concluding that the legislative history is inconclusive and that "what little legislative history there is on the point . . . tends to undercut appellants' theory," the Court has, in my opinion, taken only a superficial look into the history of the statute and has ignored the intent of Congress in various sections of the AFDC legislation as interpreted by this Court in prior cases.

B. I begin by considering the impact of § 402 (a) (23) of the Social Security Act of 1935, as amended, 81 Stat. 898, 42 U. S. C. § 602 (a) (23), on appellants' argument. That section provides that

.....

Consideration of this section must, of course, begin with Rosado v. Wyman, supra, where we examined the derivation of this section in great detail.

The relevant facts in Rosado are concisely stated in 397 U. S., at 416. New York State had changed its AFDC program so that it no longer determined need on an individualized basis, but instead substituted a system fixing maximum family allowances based on the number of individuals per family. The result was a drastic reduction in overall payments. New York State welfare recipients brought the suit in Rosado, claiming that by changing its AFDC system from an individualized-grant program to a maximum-grant program, New York had violated § 402 (a) (23).

Despite our recognition that "[t]he background of § 402 (a) (23) reveals little except that we have before us a child born of the silent union of legislative compromise," 397 U. S., at 412, we determined to discover what Congress had in mind in adding the section to the pre-existing AFDC legislation. We concluded that two general purposes could be ascribed to the section:

These conclusions led us to reject the holding of the District Court, 304 F.Supp. 1354, 1377, that Congress intended to prevent any reduction whatever in AFDC payments, and to reject the argument of the welfare recipients that if payments could be reduced § 402 (a) (23) would be meaningless. We decided that "a State may, after recomputing its standard of need, pare down payments to accommodate 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." 397 U. S., at 413 (emphasis in original). Far from emasculating the statute, our reading recognized that the statute had at least three specific salutary effects, and that these were the effects that Congress intended in enacting the legislation:

Thus, it is clear that we based our decision in Rosado, a decision that interpreted § 402 (a) (23) to permit a decrease in actual AFDC payments, largely on the conclusion that Congress wanted, not to bar decreases, but to accomplish other objectives. The fact is that the Court today undermines each of those objectives and destroys the premise on which Rosado was decided.

One specific congressional goal we saw in § 402 (a) (23) was that "[r]ecalculation of need may serve to render eligible for benefits families which may appear under unadjusted standards marginally to have attained self-sufficiency, but which in fact are unable to subsist at the present cost of living." Memorandum for the United States as Amicus Curiae in Rosado v. Wyman, No. 540, O. T. 1969, p. 8. In other words, we read the section as expressing Congress' willingness to permit reductions in actual payments in return for the addition of more families to the rolls of AFDC recipients. Accord, Lampton v. Bonin, 304 F.Supp. 1384 (ED La. 1969), vacated and remanded for reconsideration in light of Rosado, 397 U.S. 663 (1970); Alvarado v. Schmidt, 317 F.Supp. 1027 (WD Wis. 1970). As I have pointed out above, the Texas system limits the number of AFDC recipients and eliminates marginal cases. This is directly contrary to the intent of Congress as we saw it in Rosado.

A second legislative aim that we saw in the section was to force States to realize the political consequences of reducing welfare payments. It must be clear that the Texas system of administering AFDC payments effectively undermines this aim by enabling the State to maintain a constant percentage reduction factor so that the system on its face appears to contain no reductions in payments. Welfare reductions are surreptitiously accomplished by eliminating those persons who have marginal income from eligibility for AFDC payments. While the congressional intent may not be totally emasculated by this system, it is certainly not well served.

The third and final purpose that we found that Congress had specifically in mind in enacting § 402 (a) (23) was to provide an incentive to States to abandon a flat "maximum" system. Even though Texas does not now use such a system, the Court's approval of the system that Texas does use will effectively remove the incentive from the statute. A State that uses a flat maximum system was required by § 402 (a) (23) to adjust the maximums upward to reflect a rise in the cost of living. Since a State that uses a percentage-reduction system may avoid the strains cost-of-living adjustments place on the budget simply by lowering the percentage that it chooses to pay, the statute encouraged abandonment of flat maximums in favor of the more equitable percentage reductions. The Court undermines the incentive by offering States a way to circumvent the cost-of-living adjustments under the flat maximum system. In order to maintain the maximums without increasing expenditures, States could, under the Court's opinion, begin to use the maximum to determine AFDC eligibility rather than the standard of need. The result of this approach would be to reduce the number of persons eligible for assistance and to reduce the grants of anyone with any outside income. Rather than serve as an incentive to States to change to a percentage-reduction system, as Congress intended, § 402 (a) (23) may now be a powerful incentive to States to maintain or revert to maximum grants.

The manner in which the incentive that Rosado saw in § 402 (a) (23) is stifled can be illustrated by another look at the family having an income of $100 and a need of $200. Footnote 6 of the Court's opinion demonstrates that under the Texas percentage-reduction system, even if the family had no income, the maximum amount of aid that the family could obtain would be $150. Let us assume that Texas maintained a maximum grant system and that prior to the enactment of § 402 (a) (23), the maximum grant for a family with $200 need was $100. We assumed in Rosado that the following computation would be made.

    Need ....................................... $200
    Income ..................................... $100
                                                 _____

    Unmet Need ................................. $100
    Maximum Grant .............................. $100
                                                 _____

    Total Family Funds ......................... $200

Section 402 (a) (23) required an increase in the standard of need and the level of maximum grants to reflect the rise in the cost of living. Assuming that a 20% increase was mandated by the rise in living costs, it is obvious that if the number of families remained stable and if income were stable, the costs of AFDC to the State would increase by 20%. There was an incentive to change to a percentage-reduction system to avoid this. Until recently, no one thought that the State could change to the following system in order to reflect the rise in the cost of living:

    New Need ................................... $240
                                                 _____

    New Maximum Grant .......................... $120
    Family Income .............................. $100
                                                 _____

    State Aid .................................. $ 20

To state it more simply, the maximum grant is similar to, and designed to serve the same purposes as, the percentage-reduction factor. If the percentage-reduction factor can be applied to need before income is subtracted, it is impossible to see why income could not be set off against maximum grants. True, Texas did not choose this alternative, but it is available under today's decision. A State can, by changing the manner in which it sets off income, absorb an increase in maximums and end up paying less. Where is the incentive now to adopt percentage-reduction systems?

This illustration is much more than mere speculation as to what might happen under today's decision. The illustration represents what at least one State—California —has already done, or tried to do. Only very recently, the California Supreme Court struck down the State's AFDC scheme for noncompliance with the federal statute. Villa v. Hall, 6 Cal.3d 227, 490 P.2d 1148 (1971).

The California Supreme Court, having been referred to the District Court opinion in the instant case as support for California's system, took the position that neither the California nor the Texas system could stand in light of Rosado. I agree. Indeed, the United States in its Memorandum as Amicus Curiae in this case (p. 5) concedes that if Rosado represents "a binding construction of the Act, appellants are thus entitled to prevail." The Government proceeds to argue that the question presented here was not before us in Rosado. Ibid. I must agree with appellants that the Government's argument is disingenuous, at best. See Brief for Appellants 80. The question of what § 402 (a) (23) means was most certainly before us in Rosado. It was, in fact, all that was before us. In that case we rejected the broad construction that the District Court had given the section, but we endeavored as best we could to extract some meaning from its muddled history. The United States seeks here to have us do what we explicitly said we would not do in Rosado, i. e., interpret the section in such a way that it is nothing more than a "meaningless exercise in `bookkeeping.' " 397 U. S., at 413. If we were not making a "binding construction" of the statute in Rosado, it is impossible for me to ascertain what we were doing. Hence, I agree with the Government that appellants are entitled to prevail.

Surprisingly enough, the Court makes even shorter shrift of Rosado than does the Government. In a footnote, the Court states that widened eligibility and the other effects that Rosado said were intended by Congress when it enacted § 402 (a) (23) were merely possible effects of the statute, not necessary ones. I submit that this cavalier treatment of Rosado is completely unwarranted. Rosado was not an easy case. The absence of a clear legislative history forced us to examine the "muted strains" of the congressional voice and to struggle to "discern the theme in the cacophony of political understanding." 397 U. S., at 412. Unlike the Court in this case, which simply looks to see if the legislative history is distorted enough to be ignored, the Court in Rosado carefully scrutinized every aspect of the history in order to perceive the congressional intent. That was a difficult task, but not an impossible one. The balance that we saw Congress striking in reducing payments while increasing eligibility has already been described. We relied on this balance to decide Rosado. We were not merely speculating as to the intent of Congress; we were holding that there was a specific intent that was binding in that case. That decision, in my view, is also binding here. This is my first disagreement with the majority.

C. The second provision in the AFDC legislation that I believe is relevant is § 402 (a) (8) of the Social Security Act, as amended, 81 Stat. 881, 42 U. S. C. § 602 (a) (8), which was added to the AFDC statute along with § 402 (a) (23) in 1968. The purpose of this section is to encourage AFDC recipients to seek private employment and to end their need for public assistance. H. R. Rep. No. 544, 90th Cong., 1st Sess. (1967); S. Rep. No. 744, 90th Cong., 1st Sess. (1967). To accomplish this objective the statute provides that all of the earned income of each dependent child receiving AFDC aid who is a full-or part-time student, and a portion of the earned income of certain other relatives, will be disregarded in the State's determination of need. We only recently had occasion to consider the effect of this provision in Engelman v. Amos, 404 U.S. 23 (1971).

In Engelman we considered a New Jersey scheme for administering AFDC funds that established income ceilings for families. When the families' incomes exceeded the ceilings they no longer were eligible for AFDC aid. The District Court analogized Engelman to Rosado v. Wyman, supra, and determined that the State's system was inconsistent with the federal Act. 333 F.Supp. 1109. The District Court recognized that the 1968 amendments to the AFDC legislation were designed to increase eligibility for AFDC aid, not to decrease it. Because the District Court viewed § 402 (a) (8) as requiring a State to disregard certain kinds of income in determining eligibility for aid, the District Court struck down the New Jersey scheme, in effect holding that New Jersey could not evade the income disregard by imposing an income ceiling not contemplated by Congress. Families that exceeded the State's income ceilings were still entitled to AFDC aid so long as their income, excluding income covered by § 402 (a) (8), did not exceed the State's standard of need. The effect of the decision was to increase the class of persons eligible for AFDC aid. We affirmed the decision without even hearing argument.

Both "the New Jersey and the Texas provisions . . . appear to have been animated by the same desire . . . ." Memorandum for the United States as Amicus Curiae 11. Both seek to limit the number of AFDC recipients, and both violate the federal statute. Indeed, the very purpose of § 402 (a) (8)—to encourage people to work by permitting them to continue to draw AFDC funds—shows that Congress wanted as many needy people as possible to be part of the program.

The Texas scheme certainly does not violate § 402 (a) (8) in the way that the New Jersey scheme did, for as far as we know, Texas excludes income as required by the statute when computing eligibility. But, as the opinion of the Court indicates, the Texas system has a fault not found in New Jersey: i. e., Texas discourages recipients from earning outside income. This is why I believe that Texas violates the spirit of the federal statute.

It might be argued that Congress only sought to encourage certain AFDC recipients to earn income and only in a certain amount—the persons and amounts specified in § 402 (a) (8). This argument might be persuasive but for one fact—Congress never had any idea that a State would attempt to employ a system such as that used by Texas. Nowhere in the legislative history is there any mention of such a system. See, e. g., House Committee on Ways and Means, Section-By-Section Analysis of H. R. 5710, 90th Cong., 1st Sess. (Comm. Print 1967). Congress was, in fact, informed by HEW that a different standard from that used by Texas was required. See Hearings on H. R. 12080 before the Senate Committee on Finance, 90th Cong., 1st Sess., pt. 1, pp. 255-265 (testimony of Wilbur Cohen). Until very recently, every indication by HEW was that the Texas system would be unlawful. In light of the state of ignorance in which Congress found itself, it is not surprising that there is no specific rejection of the Texas system in the 1968 amendments. But § 402 (a) (8) and everything in the legislative history certainly indicate that Congress had a strong desire to encourage AFDC recipients to work. Because the Texas program is inconsistent with this desire, I believe it is illegal.

This is the second reason for my disagreement with the Court.

D. Another section of the statute that must be examined is § 402 (a) (10) of the Social Security Act, 64 Stat. 550, as amended, 42 U. S. C. § 602 (a) (10), which requires that a state AFDC plan shall

The Court states that the primary purpose of this section was to outlaw the use of waiting lists as a means of minimizing a State's welfare expenditures. There is clearly support for this view, as the Court noted in Dandridge v. Williams, 397 U.S. 471, 481 n. 12 (1970). Before the Court in Dandridge was the question whether maximum-grant limitations were inconsistent with the federal statute. The Court upheld the maximums, but said in the course of so doing: "So long as some aid is provided to all eligible families and all eligible children, the statute itself is not violated." Id., at 481. This is plainly dictum, but I believe that it is well-considered dictum that should be followed in this case.

It must be remembered that Dandridge and Rosado were decided on the same day. Thus, the Court assumed in Dandridge that the 1968 amendments to the AFDC legislation expanded the list of eligible recipients in the manner suggested in Rosado. The Court was also aware in Dandridge that § 402 (a) (7) of the Social Security Act, as amended, 53 Stat. 1379, 42 U. S. C. § 602 (a) (7), had been part of the AFDC statute since 1939. That section provides that

The Court assumed, therefore, that in offering aid a State would first set a standard of need and then examine the income level of applicants for aid. Anyone whose income was less than the standard of need would be eligible for assistance, or so the Court assumed. Dandridge, of course, established that the aid that might be forthcoming did not have to equal need and that large families could get proportionately less aid than small families. Just as in Rosado, the Court in Dandridge viewed the intent of Congress to be to aid as many needy people as possible, rather than to offer as much aid as possible to a lesser number of people. In light of this, I believe that today's decision violates the spirit of Dandridge, as well as the holding of Rosado.

Moreover, in my view, § 402 (a) (7) tells the States how to compute eligibility, and that section does not allow for the Texas scheme. Despite the position of the Government in this case, I find support for my reading of § 402 (a) (7) in HEW's own regulations, especially 45 CFR §§ 233.20 (a) (2), 233.20 (a) (3) (ii), which indicate to me that income is to be subtracted from the standard of need before any determination is made as to how much aid the State will give.

Because I believe the Texas system violates § 402 (a) (7), it seems to me that eligible persons are being denied aid in violation of § 402 (a) (10), which requires that aid be furnished to all eligible persons promptly. For me, this case is no different from King v. Smith, 392 U.S. 309 (1968) (striking down substitute-father regulation) or Townsend v. Swank, 404 U.S. 282 (1971) (striking down restriction on receipt of aid by college students). The state procedure denies eligible persons aid, and, regardless of the State's purposes, the procedure cannot stand in conflict with the federal statute.

I disagree with the Court a third time.

E. The last portion of the federal statute that I believe should be considered is that portion dealing with the social services that are available to AFDC recipients. See, e. g., 42 U. S. C. §§ 602 (a) (14), (15) (assistance in family planning and child-welfare services; assistance in entering the work force and reducing the incidence of births out of wedlock); 42 U. S. C. §§ 602 (a) (19), 632 (employment training programs); 42 U. S. C. § 1396a (a) (10) (medical assistance). Congress keyed all of these provisions to persons or families that were receiving aid. By limiting the number of such persons and families receiving aid, Texas has also limited the availability of these social services. At least one other court has concluded that

We suggested the same thing in Rosado, 397 U. S., at 413. While the Court recognizes that the Texas system deprives persons with an "unmet need" of an opportunity to utilize these services (n. 10) and thus relegates these persons to perpetual dependence on welfare, the realization is apparently a source of no concern. But it was a source of tremendous concern to Congress. The value of medical assistance alone to an average Texas AFDC family is in the range of $50-$60 per month. Memorandum for the United States as Amicus Curiae 7 n. 5. Since needy families are rendered more needy by Texas' system, their ability to escape the confines of the welfare rolls is substantially impaired. At the same time, the goals of Congress as described in the preceding quotation are also impaired. There is no reason, nor any justification, for reading the statute this way.

Since I believe that Congress intended that as many needy persons as possible be permitted to avail themselves of the various services provided or improved in the 1968 amendments, I again disagree with the conclusions of the Court.

F. In concluding my analysis of this aspect of Texas' percentage-reduction system, I add one final note. Thus far I have confined myself to examining the specific provisions of the AFDC legislation. In attempting to focus on each section individually in order to determine its role in the statutory scheme, something of the general flavor of the overall legislation is undoubtedly lost. That flavor, it seems to me, is to assist needy families to maintain strong family bonds and to assist needy individuals to realize their potential as unique human beings by providing them with the basic necessities of life, along with incentives and training to encourage them to work to help themselves. The Texas system negates the salutary aspects of the legislation by deterring the needy from working, by depriving the needy of social services, and by excluding some needy from any AFDC aid whatsoever. There is no conceivable reason to permit Texas to subvert the aims of Congress in this way.

II

Appellants also challenge the percentage-reduction figure itself. It is agreed that Texas has established an identical standard of need for the four social welfare programs that it administers—Old Age Assistance (OAA), Aid to the Blind (AB), Aid for the Permanently and Totally Disabled (APTD), and AFDC. But Texas provides 100% of recognized need to the aged and 95% to the disabled and the blind, while it provides only 75% to AFDC recipients. It is this disparity to which appellants also object.

A. Appellants base their primary attack on the Fourteenth Amendment; they argue that the percentage distinctions between the other welfare programs and AFDC reflect a racially discriminatory motive on the part of Texas officials. Thus, they argue that there is a violation of the Equal Protection Clause. I believe that it is unnecessary to reach the constitutional issue that appellants raise, and, therefore, I offer no opinion on its ultimate merits. I do wish to make it clear, however, that I do not subscribe in any way to the manner in which the Court treats the issue.

If I were to face this question, I would certainly have more difficulty with it than either the District Court had or than this Court seems to have. The record contains numerous statements by state officials to the effect that AFDC is funded at a lower level than the other programs because it is not a politically popular program. There is also evidence of a stigma that seemingly attaches to AFDC recipients and no others. This Court noted in King v. Smith, 392 U. S., at 322, that AFDC recipients were often frowned upon by the community. The evidence also shows that 87% of the AFDC recipients in Texas are either Negro or Mexican-American. Yet, both the District Court and this Court have little difficulty in concluding that the fact that AFDC is politically unpopular and the fact that AFDC recipients are disfavored by the State and its citizens, have nothing whatsoever to do with the racial makeup of the program. This conclusion is neither so apparent, nor so correct in my view.

Moreover, because I find that each one of the State's reasons for treating AFDC differently from the other programs dissolves under close scrutiny, as is demonstrated, infra, I am not at all certain who should bear the burden of proof on the question of racial discrimination. Nor am I sure that the "traditional" standard of review would govern the case as the Court holds. In Dandridge v. Williams, supra, on which the Court relies for the proposition that strict scrutiny of the State's action is not required, the Court never faced a question of possible racial discrimination. Percentages themselves are certainly not conclusive, but at some point a showing that state action has a devastating impact on the lives of minority racial groups must be relevant.

The Court reasons backwards to conclude that because appellants have not proved racial discrimination, a less strict standard of review is necessarily tolerated. In my view, the first question that must be asked is what is the standard of review and the second question is whether racial discrimination has been proved under the standard. It seems almost too plain for argument that the standard of review determines in large measure whether or not something has been proved. Whitcomb v. Chavis, 403 U.S. 124, 149 (1971); Gomillion v. Lightfoot, 364 U.S. 339 (1960).

These are all complex problems, and I do not propose to resolve any of them here. It is sufficient for me to note that I believe that the constitutional issue raised by appellants need not be reached, and that in choosing to reach it, the Court has so greatly oversimplified the issue as to distort it.

B. Appellants also challenge the distinction between programs under Title VI of the 1964 Civil Rights Act, 42 U. S. C. § 2000d:

Only last Term in Griggs v. Duke Power Co., 401 U.S. 424 (1971), we had occasion to strike down under Title VII of the 1964 Act, 42 U. S. C. § 2000e, employment practices that had a particularly harsh impact on one minority racial group and that could not be justified by business necessity. We indicated in that case that "good intent or absence of discriminatory intent does not redeem employment procedures or testing mechanisms that operate as `built-in headwinds' for minority groups." Id., at 432. We said, in fact, that "Congress directed the thrust of the Act to the consequences of employment practices, not simply the motivation." Ibid. (emphasis in original). That decision even placed the burden on the employer "of showing that any given requirement must have a manifest relationship to the employment in question." Ibid.

There has been a paucity of litigation under Title VI, and I am not prepared at this point to say whether or not a similar analysis to that used in Griggs should be used in Title VI cases. This is a question of first impression in this Court, and I do not think we have to reach it in this case. I include this section only to make plain that I do not necessarily reject the argument made by appellants; I simply do not reach it.

C. This brings me to what I believe disposes of the question presented: the disparity between the various social welfare programs is not permissible under the federal statutory framework.

The four social welfare programs offered by Texas are funded in part by the Federal Government. Each program is governed by a separate statute: OAA, 42 U. S. C. § 301 et seq.; AFDC, 42 U. S. C. § 601 et seq.; AB, 42 U. S. C. § 1201 et seq.; APTD, 42 U. S. C. § 1351 et seq. No State is compelled to participate in any program, and any State that wants to participate can choose to do so in one, several, or all of the programs.

There is no doubt that States are free to choose whether or not to participate in these programs, and it is also clear that each State has considerable freedom to allocate what it wants to one or more programs by establishing different standards of need to compute eligibility for aid. King v. Smith, 392 U. S., at 318-319. It is also true, however, that the basic aims of the four programs are identical. Indeed, when Congress first enacted the programs in 1935, it viewed them all as necessary to provide aid to families unable to obtain income from private employment. The beneficiaries of the various programs shared the basic characteristics of need and dependence. H. R. Rep. No. 615, 74th Cong., 1st Sess., 3. While the programs as they now exist go well beyond merely furnishing financial assistance as they did originally, they still maintain similar goals.

Moreover, all four programs were simultaneously amended in 1956 to provide for social and rehabilitative services to enable all needy individuals to attain the maximum economic and personal independence of which they were capable. Each program now requires a State to describe, in its plan for each social welfare program it administers, the services it offers to accomplish this objective. See 42 U. S. C. §§ 302 (a) (11); 602 (a) (14); 1202 (a) (12); 1352 (a) (11).

Congress has given the States authority to set different standards of need for different programs. But where, as here, the State concludes that the standard of need is the same for recipients of aid under the four distinct statutes, it is my opinion that Congress required that the State treat all recipients equally with respect to actual aid. In other words, as I read the federal statutes, they are designed to accomplish the same objectives, albeit for persons disadvantaged by different circumstances.

States clearly have the freedom to make a bona fide determination that blind persons have a greater need than dependent children, that adults have a higher standard of need than children, that the aged have more need than the blind, and so forth.

But, in this case, Texas made an independent determination of need, and it determined that the need of all recipients was equal. In this circumstance, I find nothing in the federal statute to enable a State to favor one group of recipients by satisfying more of its need, while at the same time denying an equally great need of another group. The purposes and objectives of the statutes are the same, those eligible for aid are suffering equally, and Congress intended that once a State chose to participate in the programs similarly situated persons would be treated similarly.

Everything in this record indicates that the recipients of the various forms of aid are identically situated. Although the District Court accepted the State's contentions that there are differences between AFDC children and other recipients which warranted different treatment under the federal statutes, I find each of the reasons offered totally unpersuasive.

First, Texas argues that AFDC children can be employed, whereas recipients of other benefits cannot be. Assuming arguendo that this is true, it is an argument that falls of its own weight. Whatever income the children earn is subtracted from need, or it is excluded from consideration under § 402 (a) (8) to encourage self-help. Thus, income is already reflected in the computation of payments, or it is excluded in order that a specific legislative goal may be furthered. Thus, income is irrelevant in any explanation of the differences between the percentage reductions applied to the various programs. It should also be noted that a recipient's income is also taken into consideration in programs other than AFDC. See 42 U. S. C. §§ 302 (a) (10) (A); 1202 (a) (8); 1352 (a) (8).

Second, the State maintains that AFDC families can secure help from legally responsible relatives more easily than recipients under other programs. Assuming again for purposes of discussion that this is true, it should be plain that any support from any relatives is subtracted from the State's grant. Moreover, appellants properly point out that recipients of aid in non-AFDC programs often have a source of aid unavailable to AFDC recipients —the federal old age insurance, 42 U. S. C. § 201 et seq. Thus, there is no substance to this argument.

Third, Texas points to the likelihood of future employment for AFDC recipients, a likelihood that it says is nonexistent for older persons and others who receive aid. Federal law provides that a State may only consider income that is currently available in allocating funds. 45 CFR § 233.20 (a) (3) (ii). This contention is therefore irrelevant.

The State makes only two other arguments. One has already been rejected. Texas urges that the purposes of the federal programs differ, but the history belies this contention. The other is that the numbers of AFDC recipients is rising and this program should therefore bear the burden of monetary limitations. The obvious problem with this argument is that one fundamental purpose of AFDC aid is to enable people to escape the welfare rolls. But, under the Texas system, the aid is presently insufficient, people are unable to escape from dependency, and the rolls become larger. Had Texas not funded AFDC at a lower level than other programs, it is possible that the number of recipients would not have grown so large. The State's argument is a self-fulfilling prophecy on which it cannot rely to penalize AFDC recipients. Furthermore, there is nothing in the federal legislation to indicate that aid is to be reduced in a program merely because the number of beneficiaries of that program increases at a more rapid rate than in other programs. On the contrary, Congress has indicated that increased eligibility for AFDC is desirable, see 42 U. S. C. § 602 (a) (23); Rosado v. Wyman, supra. It would be extreme irony if AFDC recipients were penalized by a State because their numbers grew in accordance with congressional intent.

The conclusion that I draw from the statutes is that Congress intended equal treatment for all persons similarly situated. Congress left to the States the determination of who was similarly situated by permitting States to determine levels of need. Since Texas has decided that AFDC recipients have precisely the same need as recipients of other social welfare benefits, it is my opinion that the federal legislation requires equal treatment for all.

This conclusion finds support in the legislative history of the 1950 amendments to the social welfare legislation. In those amendments Congress made clear its intent to put AFDC recipients on a par with recipients of other welfare aid.

Congress recognized that "families with dependent children need as much in assistance payments as do aged and blind persons." Id., at 106. It concluded that sound national policy was "for the States to provide payments for aid to dependent children comparable to those for the needy aged and blind." Ibid. It is evident that Congress rejected the notion that where AFDC recipients had the same need as other welfare beneficiaries, they should get less money. As Senator Benton said on the floor of the Senate:

In the 1950 amendments, Congress increased the federal funding of AFDC so that its beneficiaries would receive treatment equivalent to that received by beneficiaries of the other federal-state social welfare legislation. Where the needs of the people receiving aid under the various programs differed, Congress recognized that the amount of aid forthcoming should also differ. But where need was determined by the State to be equal for all recipients, Congress intended that all should receive an equal amount of aid. S. Doc. No. 208, 80th Cong., 2d Sess., 108. There is absolutely no indication in any subsequent congressional action that the intent of Congress has changed.

Accordingly, I would reverse the judgment of the District Court and remand the case for formulation of relief consistent with this opinion.