dissenting:
Though the Supreme Court has not proscribed all preferences, its decisions have recognized the dangers inherent in the classification of American citizens on the basis of their race. To protect against such dangers, the Court has directed that the remedy of racial preferences be dispensed only under strict conditions. Here the majority simply disregards the need for such conditions. It abandons the caution required by Supreme Court precedent and issues what amounts to an invitation to the uninhibited and ill-considered enactment of racial distinctions in local ordinances across the land. I cannot support such encouragement — either by this Court or by local authorities — of classifications shown throughout the history of this country to have the frequently-realized potential for social division and individual injustice.
I do not question for an instant the desirability of increased participation of minority business enterprises in the field of public contracts. Hasty formulas, however, only replicate the discrimination they purport to remedy and balkanize the body politic along lines of national origin and race. The program adopted by the City of Richmond runs afoul of the boundaries within which racial preferences may be pursued. Its set-aside ordinance is defective in at least two respects.
First, the city is simply unauthorized under Virginia law to enact an ordinance of this type. The sole potential source of state authority — allowing municipalities to enact procurement policies consistent with “competitive principles” — is directly contrary to the existence of a power to enact racial set-asides. It stretches the imagination to conclude that a dispensation of preferment on a basis wholly unrelated to performance or cost is consistent with competitive principles. Moreover, within the context of competition, this ordinance establishes an especially egregious regime of racial discrimination. Unlike other set-asides, it does not require minority subcontractor participation in all contracts, but draws further racial classifications by requiring only non-minority prime contractors to abide by its strictures.
Second, the factual findings made by the Richmond city council are woefully inadequate, as a matter of federal law, to ensure that the ordinance does more than “merely express[ ] a ... desire to prefer one racial or ethnic group over another.” Fullilove v. Klutznick, 448 U.S. 448, 497, 100 S.Ct. 2758, 2784, 65 L.Ed.2d 902 (Powell, J. concurring). The findings made here pay scarcely more than lip service to the indispensable preconditions for racial preferences by government. They neither support a conclusion of discrimination nor suggest the boundaries for a remedy were the need properly established. If we uphold a racial preference built on a foundation so infirm, we have in effect dispensed with the requirement for any findings at all. Local authorities need only recite an approved purpose and their racial preference will stand. I would have thought it obvious that even the best intentions of lawmakers, standing alone, are inadequate to ensure that preferences do not, in fact, diminish the dignity that resides within all individuals regardless of their race.
I
It is clear that Virginia law limits the powers of local authorities to enact racial set-asides in public contracts. This Court is duty-bound to consider and abide by those limits. See Schmidt v. Oakland Unified School District, 457 U.S. 594, 102 S.Ct. 2612, 73 L.Ed.2d 245 (1982). We do not write on a clean slate of state law, for Virginia has made clear both the severity of its limitations on municipal authority and the centrality of its concern for competition in public contracts. The putative grant of local authority here is the power *199to enact procedures of procurement “based on competitive principles.” Va.Code § 11.-35(D). The set-aside, however, introduces into the procurement process a factor wholly irrelevant to competitive concerns — the racial status of those who own or control a particular business. To equate, as does the majority, the race-neutrality of competitive contracting with the race-consciousness of numerical set-asides is to disregard principles of competition and flout settled Virginia law.
The majority misapprehends the relationship between state and local government in Virginia. The Dillon Rule limits the powers of local governing bodies to “those powers expressly granted, those necessarily or fairly implied therefrom, and those that are essential and indispensible.” Board of Supervisors v. Horne, 216 Va. 113, 117, 215 S.E.2d 453, 455 (1975). See 1 J. Dillon, Commentaries on the Law of Municipal Corporations, § 237 (5th ed. 1911). Under this rule, the state legislature is the source of all municipal authority, and its grants of authority are to be narrowly construed. Despite the dilution of the Dillon Rule in other jurisdictions, Virginia continues to follow it closely. See County Board v. Brown, 229 Va. 341, 329 S.E.2d 468 (1985). The General Assembly refused to follow the recommendation of the Commission on Constitutional Revision that the rule be reversed. Instead, the Assembly deleted the proposed repeal of the Dillon Rule from the revised state constitution. See Home, 215 S.E.2d at 456; Spain, The General Assembly and Local Government: Legislating a Constitution 1969-70, 8 U.Rich.L.Rev. 387, 403-04 (1974).
The force of the Dillon Rule in Virginia is evident from the strictness with which it is applied. Absent an express grant of the specific power in question, the Supreme Court of Virginia rarely upholds local authority to exercise the power. See, e.g., Brown, 329 S.E.2d 468 (authority to lease “unused” county land does not allow locality to lease parking lot to a developer); Tabler v. Board of Supervisors, 221 Va. 200, 269 S.E.2d 358 (1980) (authority to regulate trash does not allow locality to require deposits on disposable containers); Home, 215 S.E.2d 453 (authority to require subdivision plat approval does not allow locality to suspend filing of applications for such approval).
The Supreme Court implies local power only where an express power would be rendered ineffective absent the implication. See, e.g., Gordon v. Board of Supervisors, 207 Va. 827, 153 S.E.2d 270 (1967) (authority to create commission to develop airport necessarily includes power to lend money to commission because without such power, commission would be rendered ineffective); Light v. City of Danville, 168 Va. 181, 190 S.E. 276 (1937) (authority to build dam necessarily includes power to condemn land for that purpose). It does not imply grants of power to localities on matters of public policy as sensitive and controversial as that in issue here. See Commonwealth v. County Board, 217 Va. 558, 232 S.E.2d 30 (1977) (authority to bargain with public employee union not implied from power to enter into contracts, to hire employees, and to fix terms and conditions of employment). And it would never imply a grant of municipal power in the face of such strong statutory language to the contrary.
Against this backdrop of restrictive state law, the majority’s implication of the city’s power to enact a racial set-aside is simply untenable. The majority does not contend that the power to adopt this program is expressly granted by the Virginia Public Procurement Act. Instead, it concludes that the power is “fairly implied” from the grant of authority to base procurement policies “on competitive principles.” Far from being implied, however, the exercise of local authority here contradicts the state requirement. The power implied by the majority, consequently, finds its source in sheer judicial fiat rather than in state law.
Though the Virginia Supreme Court has not to date addressed the meaning of “competitive principles” as used in the Virginia Public Procurement Act, the term can hardly be obscure, especially in the context of *200competitive bidding for public contracts. The General Assembly has defined the term in the procedures it requires for municipalities that do not adopt their own procurement programs. The Assembly values fair and open bidding on terms of price and quality in the hope that public projects will be awarded to competent contractors at the lowest price possible.
The Assembly’s view of “competitive principles” is evident in its statement of general policies regarding procurement. The statement reflects a concern for equality of treatment for all contractors, and a distaste for arbitrary exclusions unrelated to the cost and quality of goods and services:
To the end that public bodies in the Commonwealth obtain high quality goods and services at reasonable cost, 'that all procurement procedures be conducted in a fair and impartial manner with avoidance of any impropriety or appearance of impropriety, that all qualified vendors have access to public business and that no offeror be arbitrarily or capriciously excluded, it is the intent of the General Assembly that competition be sought to the maximum feasible degree, that individual public bodies enjoy broad flexibility in fashioning details of such competition, that the rules governing contract awards be made clear in advance of the competition, that specifications reflect the procurement needs of the purchasing body rather than being drawn to favor a particular vendor, and that purchaser and vendor freely exchange information concerning what is sought to be procured and what is offered.
Va.Code § 11-35.G (emphasis added).
The specific details of the state procurement plan implement its guiding principle, the achievement of high-quality, low-cost public projects. “Competitive sealed bidding” includes at least five elements: 1) written invitations to bid including specifications and contract terms, 2) public notice of the invitation' to bid, 3) public opening and announcement of all bids received, 4) evaluation of bids based on requirements found in the invitation, and 5) award of the project to the lowest responsible and responsive bidder. Va.Code § 11-37. These requirements are, of course, familiar in the context of public contracts, for their beneficial effects are well-established and widely desired.
The state’s own program for local procurement therefore suggests a meaning of “competitive principles” consistent with the plain meaning of the term: a process in which every able entrant participates freely, with the yardsticks of performance being the lowest cost for the best quality. There is no hint that the state meant to employ the city’s conception of competition — one that apportions success on the basis of race.
The state, of course, was not blind to the need for minority business participation, but the way it chose to pursue that goal is revealing. It requires those municipalities following its procedures to include minorities in any direct solicitation of bids, Va. Code § 11-37, and has a general policy of encouraging minority participation in a way “consistent with all other provisions of this chapter.” Va.Code § 11-48. There are no numerical set-asides of contract or subcontract awards, but instead a general policy of encouraging minorities to enter the competitive process.
The Richmond plan, by contrast, is not an invitation to compete but a promise of preferment. It introduces into the otherwise cost- and quality-conscious system a factor wholly extraneous to those concerns. Able entrants in the market for subcontracts are denied the right to compete on equal terms. Results are predetermined in a way that does not comport with savings in the public fisc or quality in the public product.1
*201One cannot equate the Virginia requirement of “competitive principles” with “monopoly privileges in a ... market for a class of investors defined solely by racial characteristics.” Fullilove, 448 U.S. at 532, 100 S.Ct. at 2803 (Stevens, J., dissenting). Here, the city of Richmond has granted minority subcontractors at least oligopoly power in 30% of all subcontracts let by non-minority prime contractors. Even a limited grant of monopoly power to those not necessarily best qualified to perform a contract has adverse consequences. First, because those with such privileges have “the power to control prices or exclude competition,” United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 391, 76 S.Ct. 994, 1005, 100 L.Ed. 1264 (1956), monopoly power results in higher prices than those that would prevail under competition. See generally, F. Scherer, Industrial Market Structure and Economic Performance 14-16 (2d ed. 1980); P. Samuelson & W. Nordhaus, Economics 511-16 (12th ed. 1985). Second, the exercise of monopoly power may result in smaller output of a product than a competitive market would realize. See generally, J. Robinson, Economics of Imperfect Competition, 143-154 (2d ed. 1969). Samuelson & Nordhaus, supra at 518. These effects combine to produce a net social loss and a decline in aggregate welfare due to efficiency losses associated with monopoly power. Scherer, supra, at 459-74; Samuelson & Nordhaus, supra, at 518-19.
The facts of Croson’s case illustrate vividly the adverse effects of a set-aside program that would not occur if the city adhered to the state requirement of competitive principles. Croson in effect faced a set-aside far greater than 30%, for the supply of plumbing fixtures represented approximately 75% of the cost of the total project. Croson was obliged, therefore, to award this subcontract to a preferred subcontractor to meet the 30% set-aside. The minority subcontractor’s price for the plumbing fixtures ($96,677.14) was $6,183.29 higher than the competitive market price ($90,493.85). This would represent nearly a 7% increase in the cost of these items. If Croson had bid the job using this figure, additional overhead and bond expenditures would have brought the extra cost from using the MBE to $7,663.16. Limited public funds would thus be used on a project actually requiring a much lesser expenditure, and alternative projects — desirable and achievable in a competitive market — would be foregone or delayed.
The district court believed that “Croson’s experience ... establishes that the city considers it immaterial that an MBE provides a somewhat higher price than non-MBEs.” It rejected the argument that “any measure that may increase the cost of a contract to the city in the short run” was inconsistent with “competitive principles” by pointing to other state requirements, such as bonding, that increase costs. Set-asides, however, are quite distinct from other cost increasing measures. Racial set-asides introduce into the process a factor far removed from the business concerns that prompted a requirement for such measures as bonds that serve to insure completion of contract work. In addition, other items that increase costs are, consistent with the Dillon Rule, specifically authorized by the state. Bid bonds, performance bonds, and payment bonds, for example, are authorized by Ya.Code §§ 11.57, 11.58, 11.61 and 11.62. Absent such authorization, there would be a serious question as to the authority of municipalities to adopt these requirements. See, e.g., Tabler, 269 S.E.2d 358; Horne, 215 S.E.2d 453.2
*202Even if one were to conclude that the city could generally increase costs by taking a factor such as race into account, competitive principles would require at the very least that the city do so in a neutral manner, requiring all contractors to face the same set of increased costs for subcontracting work wherever possible. The city, however, has simply not adopted such a plan. This is evident from the fact that MBE prime contractors are exempted from the MBE subcontracting requirement. These prime contractors, therefore, enjoy a competitive advantage over their non-MBE counterparts because they can compete freely on the market for the lowest priced subcontractors, regardless of racial identity. Richmond City Code, Ch. 24.1, Art. VIII-A(A).
This reliance upon race in derogation of competitive principles is, in the last analysis, not a matter of market theory. In requiring a competitive system, the Virginia General Assembly did more than adopt an economic view. Strict procedures for competitive bidding have attempted to rid state and local government of the nemesis of kickbacks, rigged bids, and political favoritism that has bedevilled the field of public purchasing and supply almost since the advent of government. Gains in integrity have been tenuous and hard-won. To introduce a new form of favoritism invites renewed manipulations of a process which, to inspire public confidence, must remain even-handed and scrupulously fair. That this step backwards is blessed by a federal court against the aspirations of Virginia law is doubly regrettable.
II
The Richmond city council not only exceeded its authority under state law. It also failed to abide by the requirements of federal law when it enacted this set-aside. It is, of course, true that the Supreme Court has yet to articulate definitive standards for the review of such racial classifications under the equal protection clause. See, e.g., Fullilove, 448 U.S. 448, 100 S.Ct. 2758, 65 L.Ed.2d 902 (1980); University of California Regents v. Bakke, 438 U.S. 265, 98 S.Ct 2733, 57 L.Ed.2d 750 (1978). Few would deny, however, that “[a]ny preference based on racial or ethnic criteria must necessarily receive a most searching examination to make sure that it does not conflict with constitutional guarantees.” Fullilove, 448 U.S. at 491, 100 S.Ct. at 2781 (Opinion of Burger, C.J.). Here the majority abdicates that responsibility by allowing a racial preference to stand on only the barest of public records.
The need for findings prior to enactment óf a racial preference is threefold. Substantial factual findings are necessary to determine whether the purported remedy is indeed aimed at past discrimination or actually represents a bald racial preference. See Fullilove, 448 U.S. at 498, 100 S.Ct. at 2784 (Powell, J., concurring). Secondly, a court can determine that the remedy is strictly tailored to the scope of the violation “only if it is certain that the persons shaping and implementing the plan understood the nature and extent of the past discriminatory practices.” Valentine v. Smith, 654 F.2d 503, 508 (8th Cir.) cert. denied, 454 U.S. 1124, 102 S.Ct. 972, 71 L.Ed.2d 111 (1981). Thirdly, substantial findings protect against the casual use of racial distinctions and remind a public body of the seriousness of the step it is set to undertake. None of these vital purposes is served by the meager findings of the Richmond City Council in this case.
The majority’s attempt to label the findings in this case as legislative and hence all but immune from judicial scrutiny simply leaves localities unrestricted in enacting public distinctions based on race. Nothing the Supreme Court has said can be construed to give localities such license in this most sensitive of all constitutional arenas. Indeed, local bodies seeking to enact set-asides bear a heavier burden of justifica*203tion than that required of Congress in Fullilove, and therefore must make more specific findings of discrimination before adopting racial preferences. See Fullilove, 448 U.S. at 515-16 n. 14, 100 S.Ct. at 2794 n. 14. (Opinion of Powell, J.). Localities do not possess the unique remedial powers of the Congress over section 5 of the Fourteenth Amendment. See id., 448 U.S. at 483-84, 100 S.Ct. at 2777-78. (Opinion of Burger, C.J.). Accordingly, they may act only in response to well-established violations of well-defined extent. Moreover, when localities adopt set-asides, it is not always clear that a majority is acting on behalf of a minority rather than in its own favor. It is heartening that many localities are now governed by members of minority communities. At the same time, enactment of set-asides in this context may represent abuse of the political process rather than remedial action to benefit true minority interests. See generally, Ely, The Constitutionality of Reverse Racial Discrimination, 41 U.Chi.L.Rev. 723 (1974).
A.
The sparse inquiry of the Richmond City Council simply failed to adduce the significant and detailed evidence of past discrimination on which courts have conditioned use of remedial racial classifications. See Janowiak v. Corporate City of South Bend, 750 F.2d 557, 561 & 564 (7th Cir. 1984) petition for cert. filed, 53 U.S.L.W. 3896 (U.S. June 10, 1985) (No. 84-1936); South Florida Chapter of the Associated General Contractors of America, Inc. v. Metropolitan Dade County, 723 F.2d 846, 851-53 (11th Cir.), cert, denied, — U.S. -, 105 S.Ct. 220, 83 L.Ed.2d 150 (1984); Valentine, 654 F.2d at 508. Though the council adopted the set-aside plan after a public hearing, the paucity of the evidence adduced by this hearing suggests that the council performed only the most cursory inquiry into the necessity of a program for which the constitution requires searching examination.
Only seven speakers addressed the council in the most general terms during its consideration of the set-aside at the very end of a five hour council meeting. Two spoke in favor of the plan, while five opposed it. The city manager and a city councilman stated in conclusory fashion that there was discrimination in the construction industry, though the city attorney specifically denied that the city had discriminated in any individual case in the past. Several council members referred to a study purporting to show that only .67% of the city’s $124 million in construction contracts from 1978-83 had gone to minority firms. The record reveals, however, that this “study” is merely a list of city contracts without any reference whatsoever to the racial identity of any contractors. In fact, the city manager testified that overall minority participation in city contracts was 7-8%.
To uphold a racial distinction on this record makes findings nothing more than a charade to justify a pre-conceived result. The only potentially useful evidence actually before the council was the small percentage of prime construction contracts awarded by the city to minority businesses. Statistics, however, are not self-explanatory. Instead, they function as “warning signals” that “raise questions rather than settle them” and “call for further investigations into the qualitative actions and attitudes that produce the statistical profile.” United States Commission on Civil Rights, Affirmative Action in the 1980s: Dismantling the Process of Discrimination 31 (1981). See Johnson v. Transportation Agency, 748 F.2d 1308, 1319 (9th Cir.1984) (Wallace, J., concurring in part, dissenting in part). Where the need to show actual discrimination is so great, statistics cannot be invoked as a convenient “short-cut around the critical need for causal evaluation and analysis.” Johnson, 748 F.2d at 1319. Because statistics always require inquiry and explanation, “evidence of statistical disparity alone fails to prove past discrimination and cannot justify the adoption of a remedial plan that may discriminate against non-minorities.” Janowiak, 750 F.2d at 564. *204The majority, however, fails to recognize that statistical disparity suggests only the need for an inquiry, not the end of it. Though statistical disparity is consistent with a conclusion of past racial discrimination, it may also be the product of other factors. See, e.g. Milliken v. Bradley, 418 U.S. 717, 756 and n. 2, 94 S.Ct. 3112, 3133 and n. 2, 41 L.Ed.2d 1069 (1974) (Stewart, J., concurring). In particular, the availability of minority firms to perform quality work at a competitive cost must be examined. The failures of city council or prime contractors to make awards to non-existent or non-competitive concerns is more deeply regrettable than it is discriminatory. Such a finding would not support an imposition of the present competitive disadvantage on non-minority primes and subs that neither participated in nor benefitted from any pri- or malefaction.3
The cases upholding racial preferences present a dramatic contrast to the sparse evidence relied upon here. In South Forida Chapter of the Associated General Contractors of America v. Metropolitan Dade County, Florida, 552 F.Supp. 909, 925-26 (S.D.Fla. 1982), aff'd 723 F.2d 846 (11th Cir.), cert denied, — U.S.-, 105 S.Ct. 220 (1984), the municipality’s conclusion of past discriminatory practices rested not only on statistical disparity but also on evidence of “identified discrimination,” verified by independent investigations and extensive factual findings. In Bratton v. City of Detroit, 704 F.2d 878 (6th Cir.) vacated and remanded on other grounds, 712 F.2d 222 (6th Cir.1983), cert. denied, 464 U.S. 1040, 104 S.Ct. 703, 79 L.Ed.2d 168 (1984), the “inference” of intentional discrimination that arose from severe statistical disparity was supported with detailed evidence of intentional discrimination in the relevant employment context. The conclusion was further bolstered by “numerous independent studies.” See id. at 889. No such evidence was considered here; the Richmond city council relied on raw statistics. Absent further inquiry into the meaning of the numbers before it, the council’s action does not proceed from a constitutional or statutory violation and hence lacks even minimal safeguards for adoption of a racial set-aside.
B.
Detailed factual findings are also necessary to ensure that, even where an identified need exists, “the use of [racial] classifications extend[s] no further than the established need of remedying the effects of past discrimination.” Dade County, 723 F.2d at 852. Though factual findings alone do not ensure a “narrowly tailored” remedy, Fullilove, 448 U.S. at 490, 100 S.Ct. at 2781, the absence of such findings makes it impossible to limit the remedy appropriately, for there is no evidence of the scope of past discrimination at which the program is aimed. A court can determine that the remedy substantially furthers its asserted purpose only if it is certain that the persons enacting the remedy know what it was they intended to redress. See Valentine, 654 F.2d at 508. The deficiencies in the record in this regard are even more glaring than the failure to identify the discriminatory predicate.
First, the only potential evidence of past discrimination considered by the city council was unrelated to the remedy actually adopted. The evidence of statistical disparity involved the percentage of prime contracts awarded to minority businesses by the city. In response to this evidence, the council enacted a set-aside that required increased use of minorities in subcontracts. Council, however, considered absolutely no evidence on the minority subcontractor market. The district court essentially al*205lowed the council to assume discrimination against minority subcontractors from evidence of minority prime contractor under-representation. Yet it is just such offhanded reliance upon racial assumptions that the requirement of findings is designed to inhibit. Though the two situations may be related, nothing before the council demonstrated that this was so. The nexus between violation and remedy was thus left to conjecture in violation of controlling Supreme Court precedent. See Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 104 S.Ct. 2576, 2588, 81 L.Ed.2d 483 (1984); Dayton County Board of Education v. Brinkman, 433 U.S. 406, 419-20, 97 S.Ct. 2766, 2775-76, 53 L.Ed.2d 851 (1977); Milliken, 418 U.S. at 738, 94 S.Ct. at 3124; Swann, 402 U.S. at 16.
Second, the 30% set-aside goal emerges from a vacuum. Rather than a goal narrowly tailored to meet a specific need, the 30% figure is arbitrary and unsupported. No consideration was given, for example, to the number of minority firms available to perform contracts under this set-aside, despite the testimony of several witnesses before the city council that the set-aside goal was unrealistic. Though the waiver provisions may temper to some extent the quota, they do not absolve the city council of its responsibility to ensure that the goal chosen relates to demonstrated availability and need. Legislative bodies may not rectify arbitrary action through waiver provisions for which no guidelines or standards are even suggested. See Richmond City Code, Ch. 24.1, Art. VIII-A(B). The absence of any support for the percentage figure in the set-aside enhances the danger of an overbroad imposition of competitive disadvantage on non-minority contractors.
Finally, the reach of the remedy adopted here bears no relation to any geographical showing of discrimination. MBEs from all over the country may participate in the benefits of the set-aside program, yet no investigation was made as to whether any out-of-state MBEs had ever been subject to discrimination at the hands of public bodies in Richmond. Why an MBE in Pennsylvania should receive a competitive windfall in Richmond is a mystery. Surely the ordinance would have been less arbitrary if it permitted a waiver of the percentage set-aside in the event no minority subs were available in Richmond. See, e.g. Ohio Contractors Association v. Keip, 713 F.2d 167, 169 (6th Cir.1983) (set-aside limited to minority business enterprises owned and controlled by residents of Ohio). The adoption of a broad remedy for which any affirmative support is lacking* imparts a randomness to the identity of victims and beneficiaries that proper findings can do much to alleviate.
Dade County again provides a significant contrast. There, racial preferences are narrowly tailored through a series of specific steps required whenever they are to be employed. No broad remedies were enacted, instead, racial preferences are considered on a project-by-project basis. This consideration occurs in a three-stage administrative review where the whole range of race-conscious remedies — including recruitment for competitive bidding — may be considered. In no case may a set-aside be used unless three certified minority contractors are available. Finally, the entire scheme of considering race-conscious remedies is subject to annual review and assessment to determine its continuing desirability. See, Dade County, 723 F.2d at 853-854. Not a one of these ameliorating features appears in the Richmond ordinance.
IV
My intention is not to find petty fault or to demean the generous impulse from which remedial preferences spring. This question is not an easy one, and the Supreme Court has struggled to reconcile the American belief in the primacy of the individual with the need to overcome a legacy of discrimination based on race. The Court has not forbidden all remedial preferences. Rather it has flashed an amber light. Today the majority ignores that signal of caution. It approves the casual adoption of a crude numerical preference that can only impair the ideal that all stand equally be*206fore the law and postpone the day of human fellowship that transcends race.
. In contrast to the basis of authority found here, the court in South Florida Chapter of the Associated General Contractors of America v. Metropolitan Dade County, Florida, 723 F.2d 846, 852 (11th Cir.), cert. denied, — U.S.-, 105 S.Ct. 220, 83 L.Ed.2d 150 (1984), found local *201authority based on a provision that allowed the waiver of competitive bidding. No such authority exists here to justify the set-aside.
. The majority’s conjecture that the set-aside will increase competition in the long-run is unavailing. It suggests, paradoxically, that we encourage competition by use of non-competitive devices which may soon be removed, leaving those aided by such devices to confront full competition for the first time. In any event, we need not speculate about such outcomes, for the question is simply foreclosed as a matter of state law. The state has not sanctioned the abandonment of settled, commonly understood principles of competition in public contracts on *202the majority’s surmise. It has required, instead, adherence to competitive principles throughout dle operation of public procurement programs.
. Though the majority attempts to excuse the threadbare nature of the Richmond findings by an argument of “incorporation by reference,” we cannot simply assume, absent significant inquiry, that evidence relied on by Congress to adopt a national program supports the addition of another layer of racial preferences in this particular locality. If the broad congressional findings were crucial in supporting local set-asides, there would be little reason for judicial review of these set-asides at all. Courts have, however, felt a need to examine the set of findings in the particular locality. See, e.g. Dade County, 723 F.2d at 853.