United States District Court, S.D. Florida.
*1547 *1548 *1549 *1550 Charles S. Caulkins, Fisher & Phillips, Fort Lauderdale, FL, for Plaintiffs.
Robert A. Cuevas, Dade County Attorney's Office, Miami, FL, for County Defendants.
Don L. Horn, Shutts & Bowen, Miami, FL, Karen L. Stetson, Gallway, Gillman, Curtis, Vento & Horn, Miami, FL, for Intervenor-Defendant Black Business Association.
Thomas F. Pepe, Pepe & Nemire, Coral Gables, FL, for Intervenor-Defendant Allied Contractor's Association.
RYSKAMP, District Judge.
Plaintiffs are six trade associations whose members regularly perform work, either as prime contractors or subcontractors, on construction contracts awarded by Dade County, Florida. Named defendants include Dade County, Florida, its County Commissioners and the County Manager. Three parties have intervened on behalf of the defendants: the Black Business Association; the Allied Minority Contractors' Association; and the Miami branch of the National Association for *1551 the Advancement of Colored People (NAACP).
Plaintiffs are challenging Dade County's Black, Hispanic, and Women Business Enterprise Programs,[1] which provide for the use of race, ethnicity,[2] and gender-conscious measures in awarding County construction contracts, as unconstitutional under the Equal Protection Clause of the Fourteenth Amendment to the United States Constitution. Plaintiffs do not contest the manner in which the MWBE programs are administered. The Court will therefore limit its review to the programs' facial validity. Further, because plaintiffs attack the programs only with regard to construction contracting, and each program's enabling legislation contains a severability clause, the Order solely addresses the construction aspects of the program. Plaintiffs request a declaratory ruling from this Court that Dade County's MBE and WBE programs are unconstitutional and seek an injunction to prevent the County from continuing to utilize race, ethnic, and gender-conscious measures in awarding County construction contracts. A four day non-jury trial was held in this matter between December 11 and December 28, 1995, with closing arguments heard on April 11, 1996. Based on the evidence admitted at trial, and the Court's findings of fact and conclusions of law as set forth below, the Court finds that Dade County Ordinances 94-94, 94-95, and 94-96 violate the Equal Protection Clause of the U.S. Constitution. Dade County will be permanently enjoined from enforcing the race, ethnicity, and gender-conscious contract measures, as they pertain to construction projects only, contained in the aforementioned Ordinances.
Dade County first adopted a Black Business Program in 1982, by Ordinance 82-67. That program was challenged as unconstitutional by non-minority contractors. See, South Florida Chapter of Associated General Contractors of America, Inc., et al, v. Metropolitan Dade County, 552 F. Supp. 909 (S.D.Fla.1982). Under a strict scrutiny standard of review, the District Court found the "set-aside" portion of the program unconstitutional, but upheld the "goals" portion of the program.[3] In South Florida Chapter of Associated General Contractors of America, Inc., et al, v. Metropolitan Dade County, 723 F.2d 846 (11th Cir.1984), cert. denied, 469 U.S. 871, 105 S. Ct. 220, 83 L. Ed. 2d 150 (1984), the Eleventh Circuit Court of Appeals upheld the program in its entirety. Though the district court had applied a strict scrutiny standard to review the County's BBE program, the Court of Appeals reviewed the program under an analytical approach it described as "most closely akin to that set out in Chief Justice Burger's opinion in Fullilove."[4]South Fla. Chap., 723 F.2d at 852. *1552 Under the "Fullilove" standard, the Court of Appeals found the County's entire BBE program constitutional.
Since 1982, the Dade Board of County Commissioners amended the BBE program in 1984 (Ordinance 84-67), in 1992 (Ordinance 92-45), and most recently in 1994 (Ordinance 94-96). Defs' Ex. U.1. In 1994 the County adopted the Women and Hispanic Business Enterprise programs for the first time. (Ordinances 94-94 and 94-95, respectively). Id., U.2 & U.3. Because of their substantial similarity, the MBE and WBE programs will be described as one.[5]
To participate in the MWBE programs, a business must demonstrate that it does not exceed the size limits for "small business concerns" as defined by the Small Business Administration of the U.S. Department of Commerce.[6] To qualify as an MBE or a WBE the business must be owned and controlled by one or more Black individuals, Hispanic individuals, or females and must have an actual place of business in Dade County. To qualify as a joint venture, one member must be a certified MBE or WBE.
The MWBE programs apply to certain classes of County contracts for which "participation goals" have been set. For example, the County has designated Standard Industry Classification (SIC) 15, 16, and 17 for participation goals of 15% for BBEs, 19% for HBEs, and 11% for WBEs.[7] All construction contracts in excess of $25,000 funded in whole or in part by the County are covered. The County must make every reasonable effort to meet the participation goals, and may use any of the following five "contract measures" to award contracts to MWBEs in an effort to meet the participation goals.
1) Set-asides A particular contract is set aside for competition solely among one or more MWBEs. The County may waive competitive bidding on a contract, thereby "setting it aside," if there are at least three MWBE businesses available to perform the contract requirements.[8]
2) Subcontractor goals A prime contractor is required to subcontract a certain percentage of work to MWBEs. The percentage is determined on a case-by-case basis depending upon the type of subcontracting opportunities available in a particular contract and upon the availability of MWBE subcontractors to perform the work. A waiver is available to prime contractors which can demonstrate that MWBEs are not available to provide the required goods or services at a competitive price. Inability of an MWBE to obtain bonding is not considered grounds for a waiver.
3) Project goals Similar to subcontractor goals except the County creates a pool of MWBE subcontractors from which the County selects subcontractors for specified types of work under a County contract.
4) Bid preference An MWBE bid price can be reduced by as much as ten percent when comparing that bid against other *1553 bids to determine which is the lowest bid.[9] Bid preferences are given to MWBEs, joint ventures between non-minority contractors and MWBEs, and to contractors demonstrating significant utilization of MWBEs in purchasing goods and services in Dade County.
5) Selection factors Similar to a bid preference except that selection factors are used only in bids where price is one of several selection factors, such as in requests for proposals. A selection factor affords an MWBE an advantage when their proposal is evaluated.[10]
Once a contract is identified as being covered by a participation goal, it is submitted to a review committee to determine whether a contract measure should be applied. The County Commission makes the final determination concerning whether a contract measure will be applied to a contract. The decision to apply a contract measure or to award a contract under the MWBE program can be appealed to the County Manager. The County Manager's determination of the appeal is final unless the County Commission elects, at its discretion, to review the County Manager's decision.
Finally, the MWBE programs are reviewed annually by the County Commission for their efficacy. In addition, within one year of publication of the "Survey of Minority-Owned Business Enterprises" ("SMOBE" published once every five years) by the Census Bureau, the County Commission must determine whether to continue the race, ethnic, and gender-conscious measures authorized by the Ordinances.
Several non-minority contractors filed a second suit against Dade County in 1990. Capeletti Brothers, Inc., et al v. Metropolitan Dade County, 90-0678 (S.D.Fla.1990). These plaintiffs also challenged the constitutionality of the BBE program, specifically questioning whether defendants had a strong basis in evidence for the conclusion that Black contractors have suffered discrimination in the local construction industry, and whether the program was narrowly tailored. This Court held a three day bench trial in July, 1992. In August, 1992, the parties reached a settlement and stipulated to a dismissal with prejudice, before the Court rendered a final judgment.
The present case was filed in September, 1994. This action challenges the constitutionality of Dade County's MWBE programs, as they apply to the construction industry only, i.e., SIC 15, 16, and 17. Plaintiffs seek a declaratory judgment and an injunction against the enforcement of the construction industry component of the MWBE programs. Plaintiffs assert that Dade County cannot demonstrate a compelling (or, with respect to the WBE program, even an important) governmental interest in using racial, ethnic, or gender classifications as a basis upon which to award County construction contracts. Nor are the programs narrowly tailored to achieve the governmental interests identified, contend plaintiffs.
At various times during the pendency of this action, both the defendants and the intervenors have questioned whether plaintiffs have standing to bring their constitutional claims. The Supreme Court has answered this question in the affirmative. In Northeastern Fla. Chapter of the Associated Gen. Contractors of America v. City of Jacksonville, 508 U.S. 656, 665, 113 S. Ct. 2297, 2303, 124 L. Ed. 2d 586 (1993), the Court determined that construction contractors, such as plaintiffs in this case, have standing to challenge a minority preference program upon a showing that they are "able and willing to bid on contracts and that a discriminatory policy prevents [them] from doing so on an equal basis." The "injury in fact" in the context of a set-aside program "is the *1554 inability to compete on an equal footing in the bidding process, not the loss of a contract." Id. Where, as here, an MWBE program prevents otherwise qualified contractors from bidding for a given percentage of County contracts or "stacks the odds" against them through a bid preference system based solely upon their race, ethnicity, or gender, this results in a cognizable injury for purposes of an equal protection claim. Id.; see also, Contractors Association of Eastern Pennsylvania v. City of Philadelphia, 6 F.3d 990, 995 (3rd Cir.1993); Concrete Works of Colorado v. City & County of Denver, 36 F.3d 1513, 1518 (10th Cir.1994). The Supreme Court continues to find standing where the contractor plaintiff makes "an adequate showing that sometime in the relatively near future it will bid on another government contract...." Adarand Constructors, Inc. v. Pena, ___ U.S. ___, ___, 115 S. Ct. 2097, 2105, 132 L. Ed. 2d 158 (1995). Therefore, the Court finds that plaintiffs have standing to pursue their equal protection claims.
The Equal Protection Clause of the Fourteenth Amendment provides that "[n]o state shall ... deny to any person within its jurisdiction the equal protection of the laws." In its landmark decision, City of Richmond v. J.A. Croson Co., 488 U.S. 469, 487, 109 S. Ct. 706, 718, 102 L. Ed. 2d 854 (1989), the United States Supreme Court ruled on the constitutionality of government-sponsored minority preference programs. A majority of the Court agreed that the constitutionality of a state or local minority preference program must satisfy a strict scrutiny standard.
Absent searching judicial inquiry into the justification for [] race-based measures, there is simply no way of determining what classifications are "benign" or "remedial" and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.
Croson, at 493, 109 S. Ct. at 721. See also, Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 273, 106 S. Ct. 1842, 1846, 90 L. Ed. 2d 260 (1986) ("Racial and ethnic distinctions of any sort are inherently suspect and thus call for the most exacting judicial examination.")
Under a strict scrutiny analysis, race-conscious measures must be justified by a "compelling state interest," Croson, at 505, 109 S. Ct. at 727, that is "narrowly tailored" to achieve the state's compelling interest. Id. at 507-08, 109 S. Ct. at 728-29; accord, Peightal v. Metropolitan Dade County, 26 F.3d 1545, 1552 (11th Cir.1994) (Peightal II); Cone Corp. v. Hillsborough County, 908 F.2d 908, 913 (11th Cir.), cert. denied, 498 U.S. 983, 111 S. Ct. 516, 112 L. Ed. 2d 528 (1990).
Because the Croson decision is essential to the disposition of this case, it is worth taking a moment to reiterate exactly what it is, beyond the simple reaffirmation of the strict scrutiny standard for race-based classifications, that Croson says. The Court believes that the following guiding principles from Croson provide a road map by which to judge race and ethnicity-based classifications of the sort at issue here:
1. The Fourteenth Amendment guarantees the equal protection of laws to all citizens, regardless of race or ethnicity. Discrimination based upon race or ethnicity is prohibited whether invidious or benign. 488 U.S. at 493-94, 109 S.Ct. at 721-22.
2. The State, or local subdivision of the State, has the authority to eradicate the effects of private discrimination within its own legislative jurisdiction so long as this authority is exercised within the constraints of Section 1 of the Fourteenth Amendment. 488 U.S. at 491-92, 109 S.Ct. at 720-21.
3. In order to justify a program of construction set-asides or goals, the local governmental authority must show that it had become a "passive participant" in a system of racial exclusion practiced by the local construction industry. The governmental agency may not induce, encourage, or promote private persons to indulge in constitutionally forbidden prejudice. 488 U.S. at 492-93, 109 S.Ct. at 721.
4. When a local governmental agency uses a race-based measure to prohibit certain citizens the right to compete for certain *1555 public contracts, solely on the basis of race or ethnicity, the court must engage in a searching judicial inquiry into the justification for such race-based measures to determine which classifications are "benign" or "remedial" and which classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics. The purpose of this strict scrutiny is to ensure that the means used to address the compelling goal (of remedying past discrimination) "fit" so closely that there is little or no possibility that the motive for the classification was illegitimate racial prejudice or stereotype. 488 U.S. at 493, 109 S. Ct. at 721-22.
5. Local governments have a compelling interest in remedying past discrimination only after legislative or administrative findings of constitutional or statutory violations. A generalized concern regarding discrimination in society at large is not a compelling interest which will justify the use of race-based measures. 488 U.S. at 496-97, 109 S.Ct. at 723.
6. A governmental assertion that there has been past discrimination in an entire industry provides no guidance for a legislative body to determine the precise scope of the injury it seeks to remedy. Societal discrimination that has prevented minorities from following the traditional path from laborer to entrepreneur cannot justify the use of race-based measures. It is sheer speculation how many minority firms would exist absent prior past societal discrimination. 488 U.S. at 498-99, 109 S.Ct. at 724.
7. All racial classifications are suspect. A mere recitation of a legislative body that a "benign" or "remedial" purpose is in mind does not constitute a "strong basis in evidence" that the remedial plan is necessary, nor does it establish a prima facie case of constitutional violation by members of the construction industry. The local government must identify the discrimination it is seeking to remedy. 488 U.S. at 500-501, 109 S.Ct. at 725.
8. When reliance is made upon statistical disparity, and special qualifications are necessary to undertake a particular task, the relevant statistical pool must include only those minorities qualified to provide those services. These minority firms must be qualified, willing and able to provide the requested services. If the statistical analysis includes the proper pool of eligible minorities, any resulting disparity, in a proper case, may constitute prima facie proof of a pattern or practice of discrimination. 488 U.S. at 501-503, 509, 109 S. Ct. at 725-27, 730.
9. Any statistical analysis must filter out any non-social factors that may distort the final result, e.g. deficiencies in working capital, inability to meeting bonding requirements, unfamiliarity with bidding procedures, and disability caused by an inadequate track record. These problems face any new entrant into the construction industry, regardless of race. 488 U.S. at 498-99, 109 S. Ct. at 724.
10. If the governmental agency can establish a strong basis in evidence of discrimination in the construction industry, the court must determine whether the measure used to remedy prior discrimination is narrowly tailored to meet that goal. The governmental agency must show that it considered the use of race-neutral means to increase minority participation in the construction industry and rejected such means as inadequate. The use of a percentage quota is not narrowly tailored to any goal, except, perhaps, outright racial balancing. 488 U.S. at 507, 109 S.Ct. at 729.
11. When there is a significant statistical disparity between the number of qualified minority contractors willing and able to perform a particular service and the number of such contractors actually engaged by the locality and the locality's prime contractor, an inference of discriminatory exclusion could arise. Under such circumstances, the municipality could act to dismantle the closed business system by taking appropriate measures against those who discriminate on the basis of race or other illegitimate criteria. "In an extreme case, some form of narrowly tailored preference might be necessary to break down patterns of deliberate exclusion." 488 U.S. at 509, 109 S.Ct. at 730.
*1556 To summarize, then, the Croson Court declared racial quotas for state and local government contracting to be a presumptively illegal practice, reserved for extreme situations, where every constitutional "i" is dotted, and every constitutional "t" is crossed.[11]
Plaintiffs' assert that the County's WBE program also should be subjected to a strict scrutiny analysis, while the defendants contend that gender preference programs are evaluated under an intermediate scrutiny standard. The Eleventh Circuit addressed this issue in the context of a Title VII action, Ensley Branch, N.A.A.C.P. v. Seibels, 31 F.3d 1548, 1579 (1994). In the Ensley Branch, N.A.A.C.P. decision, the Eleventh Circuit rejected the argument that, based on Croson, the Supreme Court intended strict scrutiny to apply to gender-conscious programs challenged under the Equal Protection Clause, thereby implicitly reversing its application of intermediate scrutiny to gender preferences as set out in Craig v. Boren, 429 U.S. 190, 197, 97 S. Ct. 451, 457, 50 L. Ed. 2d 397 (1976).
Since Ensley, the Supreme Court handed down its decision in United States v. Virginia, ___ U.S. ___, 116 S. Ct. 2264, 135 L. Ed. 2d 735 (1996), invalidating under the Equal Protection Clause Virginia's maintenance of the single-sex Virginia Military Institution (VMI). Instead of judging the constitutionality of the VMI program under traditional intermediate scrutiny, the Court held that "[p]arties who seek to defend gender-based government action must demonstrate an `exceedingly persuasive justification for that action.'" Id. at ___, 116 S.Ct. at 2274 (citations omitted). The Court then applied this "exceedingly persuasive justification" standard in invalidating the VMI program. Justice Rehnquist concurred only in the judgment, noting that "the Court ... introduces an element of uncertainty respecting the appropriate test." Id. at ___, 116 S.Ct. at 2288. In dissent, Justice Scalia suggested that the majority had effectively adopted a strict scrutiny standard to judge the constitutionality of classifications that deny individuals opportunity on the basis of sex. Id. at ___, 116 S.Ct. at 2294. The majority, however, neither denied nor affirmed Justice Scalia's analysis.
This Court cannot say for certain whether the Supreme Court intended that the VMI decision signal a heightening in scrutiny of gender-based classifications. That issue is not dispositive of the instant case. Because the Court finds that the WBE program fails even intermediate scrutiny, it is unnecessary to decide whether the VMI decision requires the County to meet an even more difficult burden of proof. Therefore, the Court will confine its analysis of the WBE program to traditional intermediate scrutiny, noting that if the program fails this test, it would surely fail strict scrutiny as well.
Intermediate scrutiny requires that the defendants demonstrate that the gender preference is "substantially related to an important governmental objective." Contractors Ass'n v. Philadelphia, 6 F.3d at 1009. There is no clear precedent from either the Supreme Court or the Eleventh Circuit concerning what type and amount of evidence is necessary to support a public employer's gender-conscious program. The Third Circuit thoroughly analyzed the case law on this issue and fashioned a standard this Court finds persuasive. "[T]his standard requires the [county] to present probative evidence in support of its stated rationale for the gender preference, discrimination against women-owned contractors." Id. at 1010.
The parties dispute which one of them has the ultimate burden of proof on the issue of the constitutionality of Dade County's MWBE programs. While the Eleventh Circuit has not directly addressed this issue in the context of an Equal Protection claim, from its decisions, as well as those of other Circuit Courts, this Court has formulated the following approach. At the outset it must be recognized that making employment decisions on the basis of race, ethnicity, or gender is presumptively illegal. Therefore, the County bears the initial burden of showing that the race, ethnicity, and gender-conscious measures it uses to determine contract awards is not prohibited under the Fourteenth Amendment. "Because the Fourteenth Amendment only tolerates race-conscious programs that narrowly seek to remedy identified discrimination," the County must demonstrate a strong basis in evidence to support its use of race and ethnicity-conscious contract measures. Concrete Works, 36 F.3d at 1522. If the County is able to do this, then the burden shifts to the plaintiffs to rebut the County's showing. See Contractors v. Philadelphia, 6 F.3d at 1007, Howard v. McLucas, 871 F.2d 1000, 1006 (11th Cir.1989) ("[W]e emphasize that the [plaintiffs] bear the burden of proving that the affirmative action plan is unconstitutional."). "Notwithstanding the burden of initial production that rests with the municipality, the ultimate burden of proof remains with the challenging party to demonstrate the unconstitutionality of an affirmative action program." Concrete Works, 36 F.3d at 1522 (quoting Wygant, 476 U.S. at 277-78, 106 S. Ct. at 1849) (quotations omitted) and citing Johnson v. Transportation Agency, 480 U.S. 616, 626-27, 107 S. Ct. 1442, 1449, 94 L. Ed. 2d 615.
Plaintiffs have raised a second issue with regard to the admissibility of certain evidence that the Court will address here. Plaintiffs have taken pains to differentiate between "pre-enactment" and "post-enactment" evidence submitted by defendants in support of the MWBE program. Pre-enactment evidence refers to evidence developed prior to defendants enacting either the BBE, HBE, or WBE programs, and could therefore have been relied upon by the County Commission in adopting the affirmative action programs. Conversely, post-enactment evidence is that which has been developed since the County's affirmative action programs were enacted and therefore was not specifically relied upon as a rationale for the County's affirmative action efforts. The Eleventh Circuit has clearly pronounced that post-enactment evidence is properly introduced in the record and relied upon by district courts in determining the constitutionality of government affirmative action programs.
Although Croson requires that a public employer show strong evidence of discrimination when defending an affirmative action plan, the Supreme Court has never required that, before implementing affirmative action, the employer must have proved that it has discriminated. On the contrary, formal findings of discrimination need neither precede nor accompany the adoption of affirmative action.
Ensley Branch, N.A.A.C.P., 31 F.3d at 1565 (citing Wygant, 476 U.S. at 286, 106 S. Ct. at 1853; Howard, 871 F.2d at 1007; Contractors Ass'n v. Philadelphia, 6 F.3d at 1004; and Harrison & Burrowes Bridge Constructors, Inc. v. Cuomo, 981 F.2d 50, 60 (2nd Cir.1992)). The Third Circuit's perspective on this issue is also noteworthy:
Consideration of post-enactment evidence is especially appropriate here, where the principal relief sought and the only relief granted by the district court, was an injunction. Because injunctions are prospective only, it makes sense to consider all available evidence before the district court, including the post-enactment evidence, which the district court did. Although we recognize the risk of insincerity associated with post-enactment evidence, we believe the risk is minimal here because the [evidence] consists essentially of an evaluation and re-ordering of pre-enactment evidence contracts awarded to minority owned businesses in the three years preceding the Ordinance.
Contractors Ass'n, 6 F.3d at 1004.
Although post-enactment evidence may be probative of discrimination, the Court finds *1558 that much of the statistical evidence regarding BBE contracting after 1982 is severely skewed by the challenged affirmative action program. The program at once allocated more contracts to BBEs, and provided incentives for more blacks to enter the contracting business and bid on County contracts. Thus, the ordinance artificially inflated the demand for County contracts among BBEs and the number of BBEs being awarded contracts. The converging phenomena of increasing BBE demand for contracts and increasing BBE success in bidding make it difficult to determine the underlying issue whether the County was guilty of discrimination during the years in question.
Additionally, the use of data not merely organized and studied post-enactment, but actually culled from years while the ordinance was in effect, creates the following dilemma for defendants: if the data prove the existence of discrimination, then they also prove that the set-aside program is not achieving its goal of ending discrimination. If the set-aside program is ineffective, it can hardly be considered necessary to promote the County's compelling interest in ending discrimination. This reasoning may appear tautological, but it is a tautology created by defendants' principal reliance on data culled from the years 1989 through 1991, years while the ordinance was in effect.
At least as to BBEs, the Court would prefer to rely on the data from 1982, the last year prior to the enactment of the ordinance, which seems to offer the best window on the County's treatment of black contractors apart from the set-aside program.[12] However, since much of the trial centered upon statistical data culled from years while the BBE program was in effect, the Court will consider that data as well.
Requiring a state government to demonstrate a compelling interest "was designed to `smoke out' illegitimate uses of race by assuring that the legislative body is pursuing a goal important enough to warrant use of a highly suspect tool." Ensley Branch, N.A.A.C.P., et al, v. Seibels, 31 F.3d 1548, 1565-66 (11th Cir.1994) (quoting Croson, 488 U.S. at 493, 109 S. Ct. at 721). The Eleventh Circuit has determined that "a compelling state interest unquestionably exists where race-conscious relief is aimed at past and present discrimination by a state actor." Peightal II, 26 F.3d at 1552 (citing United States v. Paradise, 480 U.S. 149, 167, 107 S. Ct. 1053, 1064, 94 L. Ed. 2d 203 (1987)) (internal quotations omitted); see also, Ensley Branch, N.A.A.C.P., 31 F.3d at 1565 (Remedying past or present discrimination is widely accepted as compelling.)
Thus, the Eleventh Circuit has explained that "the true test of an affirmative action program is usually not the nature of the government's interest, but rather the adequacy of the evidence of discrimination offered to show that interest." Ensley Branch, N.A.A.C.P., 31 F.3d at 1565 (citations and internal quotations omitted). "[A] district court must make a factual determination that the public employer has a strong basis in evidence for its conclusion that racial discrimination necessitates affirmative action." Id. "The exact quantum of evidence required eludes precise definition" but "the proffered evidence must approach a prima facie case of a constitutional or statutory violation" Peightal II, 26 F.3d at 1553 (citing Croson, 488 U.S. at 500, 109 S. Ct. at 725) (internal quotations omitted). "Evidence of "gross statistical disparities" between the number of MWBEs being utilized in Dade County construction contracts and the number of MWBEs available to do the work required under the County contracts "may constitute prima facie proof of a pattern or practice of discrimination." Id. (citing Croson, 488 U.S. at 501, 109 S. Ct. at 725-26).
Dade County must therefore demonstrate that gross statistical disparities exist between the proportion of MBEs awarded *1559 County construction contracts and the proportion of MBEs in the local construction industry "willing and able to do the work" in order to justify its use of race and ethnicity-conscious contract measures. Ensley Branch, N.A.A.C.P., 31 F.3d at 1565; see also Croson, 488 U.S. at 501-02, 109 S. Ct. at 726 ("[W]here special qualifications are necessary, the relevant statistical pool for purposes of demonstrating discriminatory exclusion must be the number of minorities qualified to undertake the particular task.") "To adequately assess statistical data, there must be evidence identifying the basic qualifications" of an MBE "willing and able to do the job" and the Court must make a "determination, based upon these qualifications, of the relevant statistical pool with which to make the appropriate [statistical] comparisons." Peightal II, 26 F.3d at 1553-54 (citations omitted). The Croson Court recognized the utility of a type of statistical analysis known as a disparity index for purposes of comparing the proportion of MWBEs qualified to perform the contracts at issue with the proportion of MWBEs actually being awarded the contracts in order to demonstrate the existence of past or present discrimination in the local construction industry.[13]Croson, 488 U.S. at 501-02, 109 S. Ct. at 726. Accordingly, Dade County has conducted disparity analyses for BBEs, HBEs, and WBEs in an effort to show that the proportion of qualified MWBES being utilized in County-funded construction is significantly lower than the proportion of qualified MWBEs available in the local construction industry to do the work.
A disparity index is calculated by dividing the percentage of MWBEs participating in city contracts by the percentage of qualified MWBEs in the relevant population of local construction firms. Concrete Works, 36 F.3d at 1523 n. 10. The resulting percentage is then multiplied by 100 to yield a number between 0 and 100, with 100 indicating parity, or full participation by MWBEs, given their proportion in the total population of available, qualified contractors. Contractors Ass'n, 6 F.3d at 1005.
The number calculated via the disparity analysis, the disparity index, is then tested for its validity through the application of a standard deviation analysis.[14] The Eleventh Circuit has directed that "where the difference between the expected value and the observed number is greater than two or three standard deviations" a presumption of discriminatory conduct is raised.[15]Peightal II, 26 F.3d at 1556 (citing Hazelwood School Dist. v. United States, 433 U.S. 299, 308 n. 13, 97 S. Ct. 2736, 2742 n. 13, 53 L. Ed. 2d 768 (1977)). In order to be probative, the "numerical disparities" presented by the defendants must "result from a comparison of the number of qualified MBEs in the particular industry and geographic area with the number of MBEs actually utilized in government contracts." Cone Corp., 908 F.2d at 914. Affirmative action programs can be justified by "demonstrating gross statistical disparities between the proportion of minorities hired by the public employer and the proportion of minorities willing and able to do the work." Ensley Branch, N.A.A.C.P., 31 F.3d at 1565; see also Croson, 488 U.S. at 501, 109 *1560 S.Ct. at 726. "At a minimum, however, the government entity must consider the number of qualified MBEs in the relevant market and the percentage of total city dollars minority firms receive." Id.
Defendants have proffered several statistical analyses of County construction contracting trends that they claim provide them a strong basis in evidence to show that discrimination occurs in the local construction industry and thus supports the use of race, ethnicity, and gender-conscious measures by the County in awarding contracts.
Defendants' most compelling data, and that which they focused on most heavily at trial, are presented in "A Minority- and Women-Owned Business Discrimination Study" done for the County by MRD Consulting, Inc., and discussed at trial by Dr. Manuel Carvajal, the co-author. Defendant's Exhibit D (composite), Volumes 1-6 (hereinafter "MRD study"). The bulk of the statistical analysis contained in the MRD study focuses on Dade County procurement contracts. Possibly because construction contracts only comprise ten percent of procurement contracts, at trial the statistical evidence defendants presented focused almost exclusively on nonprocurement contracting, which was a small part of the MRD study completed after the main study. Tr. 242 34. Defendants' expert, Dr. Carvajal, testified that the results of the nonprocurement statistical analysis were the most probative of discrimination that the defendants had. As a result, the Court will focus its analysis on the nonprocurement data commensurate with defendants' evidence and testimony at trial.
The MRD study presents the nonprocurement contract data aggregated for all capital construction contracts (i.e., SIC 15, 16 and 17 together) and disaggregated by SIC category: SIC 15 general building construction; SIC 16 heavy construction other than building construction; and SIC 17 specialty trade construction, which includes electrical, plumbing, heating, ventilating, and air conditioning (HVAC), among others. The Court will focus primarily on statistical analyses of the disaggregated data because these data are more likely to reflect the realities of competition in the construction industry. Firms that build hospitals (SIC 15) do not compete for County contracts with firms that lay asphalt (SIC 16) or firms that install plumbing (SIC 17), therefore comparisons between these disparate entities would not produce a reliable portrait of County contracting trends.[16] The MRD statistical analysis evaluates County contracting trends for discrimination in the award of County contracts to BBEs, HBEs, and WBEs. The Court will discuss each in turn.
The MRD study analyzes two pools of Black-owned prime contracting firms, those BBEs that bid at least once for a County capital construction contract (SIC 15, 16, & 17) during the relevant time period (bidders), and those BBEs that were awarded at least one County capital construction contract over the relevant time period (awardees). Defendants' and Intervenors' Post Trial Proposed Findings of Fact and Conclusions of Law, 12-13 (hereinafter Defs' Prop. Findings); Defs' Ex. L-1. The MRD study looked at two distinct time periods, 1989-1991, and 1993.[17]
*1561 The proportion of BBEs that bid on at least one County construction contract within each SIC category is compared to the proportion of contract dollars awarded to BBEs within that SIC category. This is done for each SIC category and for 1989-1991 and for 1993. Comparing the proportion of BBEs to the proportion of contract dollars awarded results in a disparity index.[18] This figure is then evaluated through the use of standard deviation analyses to determine whether the figure has statistical significance, that is, whether the standard deviation is greater than two.[19]
When the BBE bidder and awardee data are disaggregated by type of construction contract: SIC 15 general building construction; SIC 16 heavy construction other than building construction; and SIC 17 specialty trade construction; and compared to the proportion of contract dollars awarded to BBEs during 1989 to 1991, the results are as follows:
BBE Bidders(%) BBE Awardees(%) Contract $(%)
SIC 15 13.8 15.0 1.8
SIC 16 5.2 3.4 0.5
SIC 17 16.2 13.5 4.8
Defs' Prop. Findings 13; Defs' Trial Exhibit L-1.
For SIC 15, the disparity between BBE bidders (13.8) and contract dollars awarded to BBEs (1.8) is 12.6, which is statistically significant at a level of three standard deviations. The disparity between BBE awardees (15.0) and contract dollars (1.8) is 11.6, also significant at three standard deviations. The disparity between BBE bidders in SIC 16 (5.2) and contract dollars awarded in SIC 16 (0.5) is 10.1, at 2.6 standard deviations of statistical significance. The disparity for awardees in SIC 16 (3.2) versus contract dollars awarded in SIC 16 (0.5) is 15.5 with statistical significance of 1.9 standard deviations. Finally, for SIC 17, the disparity between BBE bidders (16.2) and contract dollars awarded (4.8) is 29.7, with a standard deviation of 2.3, and for BBE awardees (13.5) versus contract dollars (4.8) is 35.6, at 1.8 standard deviations.
In 1993, when disaggregated, the County contracting awards looked like this:
BBE Bidders(%) BBE Awardees(%) Contract $(%)
SIC 15 17.5 24.6 7.8
SIC 16 16.6 24.1 9.9
SIC 17 21.3 20.0 14.0
Defs' Prop. Findings 13-14; Defs' Trial Exhibit L-3.
For SIC 15, the disparity between BBE bidders (17.5) and contract dollars awarded to BBEs in SIC 15 (7.8) is 44.9, which is statistically significant at a level of 1.8 standard deviations. The disparity between BBE awardees in SIC 15 (24.6) and contract dollars (7.8) is 31.9, significant at 2.6 standard deviations. The disparity between BBE bidders (16.6) and contract dollars awarded (9.9) in SIC 16 is 59.3, at 1.4 standard deviations of statistical significance, and for awardees in SIC 16 (24.1) versus contract dollars (9.9) is 40.9 with significance of 2.5 standard deviations. Finally, for SIC 17, the disparity between BBE bidders (21.3) and contract dollars awarded (14.0) is 65.6, with no statistical significance, and for BBE awardees *1562 in SIC 17 (20.0) versus contract dollars (14.0) is 69.9, with no statistical significance.
What stands out about data in this part of the MRD study is that the proportion of minority firms that bid on County contracts is not remarkably different than the proportion of minority firms awarded County contracts. For example, in 1993 in SIC categories 15 and 16, BBEs were awarded proportionately more County contracts than would be expected given their proportion in the population of construction firms bidding for those contracts.[20] This is also the case for SIC 15 in the years between 1989 and 1991.[21] Thus, in half of the SIC categories analyzed, the disparity indices are greater than one when comparing number of BBEs bidding to number of BBEs being awarded contracts, indicating no disparities exist in the proportion of contracts awarded to Black-owned firms for these categories and time periods.
The remaining three SIC categories do indicate disparities exist between the number of BBEs bidding on County contracts and the number of contracts awarded to BBEs. However, there are several factors that must be considered in determining the significance of these disparities. First, in two of the SIC categories the disparities are not substantial. In SIC 17, for the years 1989 through 1991, 16.5% of the bidders were BBEs, and they received 13.5% of the contracts awarded. This results in a disparity index of 83.3%. For 1993, BBEs comprised 21.3% of the firms bidding for contracts in SIC 17, and they received 20% of the contracts awarded, for a disparity index of approximately 94%. For SIC 16, BBEs made up 5.2% of the firms bidding on County contracts and received 3.4% of the contracts awarded. This results in a disparity index of 65.4.
In general, disparity indices of greater than 80% are not considered substantial.[22] In addition, it is not clear to the Court that the denominator used in the disparity analysis, that is the number of BBEs that bid on a Dade County contract within a certain SIC category, accurately reflects the relevant pool of willing, able and qualified minority firms. Dade County does not have a prequalification process through which the County determines whether a particular firm is qualified to do the work required under a specific contract.[23] The only pre-bid qualification the County requires a contractor to meet is that the firm be a licensed contractor able to satisfy the bonding and insurance requirements for the contract being bid upon. This, of course, does not reflect a firm's ability to perform the work required under the contract.
In three of the six SIC categories analyzed, BBEs were getting proportionately more contract awards than non-MWBEs. Of the three categories in which there were disparities between the number of BBEs bidding and the number of contracts awarded BBEs, two categories had nominal disparity indices of 83% and 94%. This does not provide a strong basis in evidence to conclude that discrimination exists in the process by which the County awards contracts, nor that there is discrimination in the local construction industry that results in BBEs being prevented from successfully bidding for County contracts.
This leaves one SIC category in which there is a substantial disparity between the number of BBE bidders and the number of contracts awarded. For SIC 16 over the years 1989 1991, an analysis of the numbers results in a disparity index of 65.4%. The relevance of this number is unclear for two reasons. First, the underlying data set is relatively small. Only about eight firms out of the 155 that bid on County contracts *1563 in SIC 16 over the period 1989-1991 were Black-owned. These eight firms were awarded approximately six contracts out of the 178 the County let over that period. To be strictly proportionate with the percentage of BBE bidders, BBEs should have been awarded nine County contracts over that time period rather than six. There are a myriad of possible reasons for this discrepancy. It could be that, though there were eight BBEs bidding, not all of these firms were actually qualified to do the work. Perhaps, given that SIC 16 is heavy construction (except for building), the BBEs bidding were too small for the typical size contract awarded for highway construction or bridge construction. Conceivably, the BBEs were simply not competitive in their bids. And, possibly, the failure to receive three contract awards could be due to discrimination, either within the County government or in the local construction industry as a whole.
The second reason the disparity in SIC 16 for 1989-1991 is suspect is because in 1993 the disparity index is positive for BBEs, at 1.45. Thus, BBEs actually received significantly more contract awards than would be expected given their proportionate share. This is not to say that the Black-owned firms did not deserve or were not qualified for the contracts they were awarded. The Court merely observes that the phenomenon whereby a firm successfully bids upon and is awarded a contract at any given time is dependent on a multitude of factors in a competitive and dynamic system, including the availability of suppliers, subcontractors, other contract commitments the firm may have, and the firm's ability to obtain the financing and bonding required to do the work. While a firm may be an unsuccessful bidder at one time for any one of these reasons, the same firm may well be successful at the next bid opportunity because of changed circumstances.
Because of the lack of convincing evidence that there is any discernable discrimination occurring in the award of contracts to BBEs, the defendants understandably have focused on the disparity between the proportion of BBE bidders and the proportion of contract dollars these firms receive. Essentially, defendants are asserting that, despite the fact that BBEs are getting their fair share of County construction contracts, the County award process is biased in a way that results in BBEs getting a disproportionate share of lower dollar contracts.
Plaintiffs contend that the disparities defendants point to in contract dollars awarded do not necessarily mean that discrimination is occurring in the award of Dade County construction contracts. Defendants' experts agree that disparities alone do not prove that discrimination is in fact occurring. Tr. 34, 40, 142. Plaintiffs' argue that the disparities depicted by the MRD study simply reflect the fact that MWBEs tend to be newer, smaller, and less experienced firms the majority having entered the construction industry in the last ten years and thus are more often successful bidders on lower dollar contracts rather than higher dollar contracts. The MRD study notes, and defendants' expert testified to at trial, MWBEs tend to be smaller and less established. Tr. 158; Defs' Ex. D, vol. 1, p. 81 & vol. 3, p. 1. The MRD study also states that the majority of non-MWBEs have been in business more than ten years. Defs' Ex. D, vol. 1, p. 65. That MWBEs are on average smaller is buttressed by Dade County census and survey data.[24] The evidence plaintiffs introduced suggests that, as of the 1987 Business Census, BBEs had 3.1 employees on average, HBEs had an average of 4.3 employees, WBEs had 6.6 employees on average, and non-MWBEs had an average of 14.1 employees. Plntfs' Ex. 67.
The tables developed pursuant to the MRD study are further evidence that MWBEs tend to be smaller in size than non-MWBEs. For example, the MRD study estimated the size of firms by determining the largest County contract award a firm had bid upon. For this purpose, the study broke down contract awards into three categories: *1564 awards under $500,000, awards between $500,000 and $2,999,999, and awards over $3 million. Of the 20 BBEs that bid for County contracts under SIC 15 general building construction between 1989-1991, 14, or 70 percent, submitted bids no larger than $500,000. The largest bid submitted for the remaining 30 percent is between $500,000 and $2,999,999. No BBEs bid on County contracts over $3 million during the three year period. Defs' Ex. A-1, Table C-42, pp. 3-4. For the same SIC category and time period, there were 77 non-MWBE bidders. Nineteen had largest bids of less that $500,000 (25%), thirty had largest bids between $500,000 $2,999,999 (39%), and twenty-eight had largest bids of over $3 million (36%). Id.
Of the eight BBEs bidding for contracts awarded under SIC 16 heavy construction (except building) the largest bid submitted for four of the eight is under $500,000 (50%). Three BBEs submitted their largest bid between $500,000 and $2,999,999 (37%). One BBE bid on an award of over $3 million (13%). Id. at 5-6. Non-MWBEs bidding for contracts in SIC 16 from 1989-1991 totaled 107 firms. Thirty non-MWBEs submitted maximum bids of under $500,000 (28%), while 32 submitted largest bids between $500,000 and $2,999,999 (30%), and 45 had maximum bids over $3 million (42%). Id.
The trend is similar for SIC 17 specialty trade construction. Of the 17 BBEs that bid on these contracts between 1989 and 1991, 13 had largest bids under $500,000 (76%), 4 had largest bids between $500,000 (24%), and there were no BBE bidders for contracts over $3 million. Id. at 7-8. Non-MWBEs comprised 53 of the firms bidding overall, 31 with a maximum bid of less than $500,000 (59%), 14 with maximum bids between $500,000 and $2,999,999 (26%), and 8 with maximum bids over $3 million (15%). Id.
The totality of this evidence clearly supports the conclusion that BBEs tend to be smaller contractors, the majority competing for contract awards of under $500,000. Conversely, non-MWBEs are almost evenly distributed over the range of County contracts awarded. It is important to note that the average capital construction contract let by Dade County is worth approximately $3 million. Defs' Ex. D, vol. 6, p. 2. In order to bid and win a large contract, it is reasonable to assume that a firm must be sufficiently large and established to achieve the financing, bonding and insurance requisites necessary to put forth a successful bid. Given this, it is likewise reasonable to conclude that larger firms may, on average, have higher dollar contract awards. Concomitantly, smaller firms would be expected to have smaller average dollar awards. If, as the evidence indicates, MWBEs tend to be, on average, smaller, and non-MWBEs tend to be larger, this could account for disparities in the average size of the County contract awarded.
The strength of this logic has not been lost on the defendants. Recognizing that the size and experience of a firm can impact both the likelihood of being awarded a contract and the dollar value of the contract awarded, defendants attempted to account for these factors by performing regression analyses. Regression analysis is a statistical model used to estimate both the nature and magnitude of the relationship that exists between a dependent variable Y and one or more independent or explanatory variables X. For example, a dependent variable could be dollar value of a contract award and an explanatory variable could be the size of the firm. A regression using these variables attempts to assess whether the size of the firm correlates to the dollar value of a contract a firm is awarded. Defs' Ex. D, vol. 3, p. 101.
As defendants' expert explained, a regression analysis evaluates whether an event is happening due to chance or random occurrences, or whether the event is occurring as a result of a systematic pattern that is not attributable to chance. Tr. 140, 159. Defendants argue that their statistical analyses show that the disparity between the proportion of BBEs awarded County contracts and the proportion of contract dollars received by BBEs is not the result of the size or experience differential between BBEs and non-MWBEs, nor is it the result of chance, rather it is the result of a pattern and practice of *1565 discrimination against BBEs in the award of Dade County contract dollars.[25]
Whether a regression analysis indeed shows that the award disparities are due to discrimination is determined by the level of statistical significance of the regression equation. The higher the level of statistical significance, the less likely an event is random. By regressing for an explanatory variable such as size of firm, defendants are attempting to show that the disparity in contract dollars awarded to BBEs is not a result of the size of the firm, rather, it is attributable to the fact that the firm is minority-owned versus non-minority owned. Tr. 159. In order for this explanation to be persuasive, the statistical significance of the regression equation must be between two and three standard deviations.[26]
Defendants' study utilizes two explanatory variables to account for firm size: total awarded value of all contracts bid on; and largest single contract awarded. Tr. 152; Defs' Ex. L-2. When controlling for firm size by regressing for total awarded value of all contracts for the years 1989-1991, the only SIC category in which the award disparity appears to be statistically significant is SIC 15, heavy building construction, at 2.5 standard deviations. Defs' Ex. L-2. The same regression for SIC 16 results in a standard deviation of 1.3. Id. For SIC 17, the regression indicates that there is no statistically significant disparity in contract dollars awarded. Id.; Tr. 161-62.
The results are substantially the same when controlling for firm size by regressing for largest contract awarded in the years 1989-1991. Again, the only SIC category in which the award disparity appears to be statistically significant is SIC 15, at 2.8 standard deviations. Defs' Ex. L-2. This time the regression for SIC 16 results in no statistically significant disparities in contract dollars awarded. Id. For SIC 17 the regression suggests a statistical significance of 1.8 standard deviations. Id.
The MRD study also presents the results of these same regression analyses run for the year 1993. Defs' Ex. L-4. None of the SIC categories analyzed for total awarded value of all contracts or for largest contract awarded demonstrate statistically significant disparities of two standard deviations or more. Id. For total awarded value of contracts bid in 1993, SIC 15 shows no statistical significance. The regressions for SIC 16 and 17 both result in 1.2 standard deviations. Id. The regression analysis for largest contract awarded in 1993 indicates that the disparities in SIC 15 are at 1.2 standard deviations, while SIC 16 and 17 demonstrate no statistically significant disparities at all.
The Court does not find this evidence particularly persuasive. The only SIC category for which there appeared a statistically significant disparity between the dollars awards going to BBEs versus dollar awards going to non-MWBEs is SIC 15 over the time period from 1989-1991.[27] By 1993, the regression analyses no longer indicates that the disparities in SIC 15 are statistically significant. The results of the regression analyses appear to be problematic for other reasons. For example, at trial defendants' expert admitted that there are instances in which their regression analyses indicate that the more experience and success a firm has had in bidding on Dade County contracts, the less likely that firm is win a County contract in the future. Tr. 231-235. This conclusion would *1566 seem counterintuitive, especially since the testimony from experts on both sides and the studies submitted by both sides exhibits a general consensus that the more experience a firm has, and the larger a firm is, the greater the number and dollar value of contracts a firm can successfully bid upon.
In addition, the Court is not convinced of the reliability of the data underlying this portion of the MRD study. Although the bulk of the statistical analysis contained in the MRD study focuses on Dade County procurement contracts, at trial defendants statistical evidence focused almost exclusively on a small part of the MRD study completed after the main study which dealt with nonprocurement contracting. Tr. 242-34. Defendants' expert testified that the results of the nonprocurement statistical analysis were the most probative of discrimination, Tr. 240, but the MRD study itself offers the following caveat regarding the reliability of the non-procurement data:
[T]he [nonprocurement data] cannot be nearly as accurate as the estimates derived from the procurement file for several reasons. First, the nonprocurement data set is a sample, whereas the procurement data set represents the universe. Second, contractor-related information gathered for the nonprocurement analysis is derived from telephone interviews; answers are subject to contractors' availability and willingness to respond, as well as the accuracy of their response ... And third, nonprocurement activities are far more hetrogeneous than procurement activities.
Defs' Ex. D, vol. 6, p. 1. Nonetheless, defendants did not utilize any of the procurement data at trial, relying instead upon the nonprocurement data as the foundation of their case, despite their acknowledged weaknesses. For these reasons, the Court finds that the statistical data presented to show discrimination in the award of prime contracts to BBEs fall short of providing a strong basis in evidence for the BBE program.
Defendants also presented disparity and regression analyses for a pool of firms described as subcontractors. The firms evaluated in this portion of the MRD study were identified as subcontracting firms because each of them has filed a subcontractors release of lien on at least one Dade County contract over the period 1991-1994.[28] The statistical analyses were performed for each SIC category and for BBEs, HBEs and WBEs. See, Defs' Tr. Ex. L-5, L-6, M-5, M-6, N-5, N-6 ("X" files).
Defendants' analysis compares the proportion of BBEs that filed a county release of lien with the proportion of sales and receipts claimed by BBEs. Defs' Tr. Ex. L-5. For SIC 15, between 1991 and 1994, 11.6% of the firms that filed a release of lien were BBEs, while the sales and receipts of these BBEs comprised 1.3% of total sales and receipts claimed by all firms which had filed a release of lien with the County. This results in a disparity index of 11, at 1.8 standard deviations, which is not statistically significant to a level of two standard deviations. The disparity index calculated for SIC 16 had no calculated statistical significance whatsoever. For SIC 17, the disparity index between BBEs and sales receipts is 54.9. Again, the statistical significance is low, at 1.3 standard deviations.
Because the size of a firm impacts the size of its sales and receipts, defendants performed regression analyses using number of employees as an explanatory variable to control for the impact of size of firm on the dependent variable sales and receipts. The regressions for all of the SIC categories were below 2 standard deviations of statistical significance. Defs' Ex. L-6.
In addition to the striking fact that none of the disparity or regression analyses resulted in a statistical significance of two standard deviations, the data relied upon do not lend themselves to developing accurate conclusions about subcontracting trends in Dade County. The first problem is that the *1567 pool of identified subcontractors being evaluated may not in fact derive their primary source of income from subcontracting. The fact that a firm has filed a subcontractors release of lien does not preclude that firm from deriving most of its income from other sources, such as prime contracting. More importantly, the sales and receipts data used in the statistical analyses come from all sources, and is not limited to sources within Dade County. A subcontractor that does the majority of its work in Broward or Palm Beach County, or perhaps outside the state of Florida, but has done one job in Dade County, would have all of its sales and receipts included in this evaluation. The Supreme Court has made it clear that an appropriate statistical evaluation must confine itself to the geographic boundaries of the jurisdiction to which the affirmative action program applies.[29]
The Croson Court found Richmond's justification for its subcontracting measures unpersuasive because, inter alia, there was no evidence of the proportion of contract dollars minorities received as subcontractors. Croson, 488 U.S. at 502, 109 S. Ct. at 726. Dade County has likewise been unable to produce evidence on this point. Because the data underlying the defendants' subcontracting analysis are inappropriate, and the statistical analyses themselves lack the requisite statistical significance, the Court finds the subcontracting study insufficiently probative to support the use of race and ethnicity-conscious measures in the form of subcontracting goals in the County's BBE program.
Defendants also prepared a series of tables that Dr. Carvajal described as a marketplace analysis. The pool of firms analyzed includes a sample of 586 contractors, out of a total of 10,462 firms which had filed a certificate of competency with Dade County as of January 1995. Tr. 168; Defs' Ex. L-7 ("W" files). Receiving a certificate of competency does not mean a firm is willing, able or qualified to perform work on County construction projects, it simply means the firm has registered as a construction contractor with the County. The sampled firms were surveyed by phone for information on the race, gender, and ethnicity of the owner and its total sales and receipts from all sources, both public and private, within Dade County and without. Tr. 171. The purpose of this study is not to demonstrate that there is discrimination in Dade County contracting which has resulted in disparities in the total sales and receipts of Black-owned firms versus non-MWBE-owned firms. Rather it is to examine whether a sample of construction firms which have registered with Dade County would, when surveyed and analyzed, exhibit disparities in the average sales and receipts of Black- versus non-MWBE-owned firms.
Given that non-MWBEs tend to be larger than BBEs, it is not surprising to the Court that the analysis showed disparities exist, though only SIC 17 exhibited a statistically significant disparity of 2 standard deviations. Defs' Ex. L-7. After running regressions using number of employees as a proxy for size, none of the SIC categories exhibited disparities of greater than 1.4 standard deviations. Defs' Ex. L-8. Absent a more robust level of statistical significance, the Court is not convinced that the difference in the average amount of sales and receipts between BBEs and non-MWBEs is attributable to anything more than the difference in average size of BBEs versus non-MWBEs firms.
The MRD study analyzes Hispanic-owned firms using the same disparity and regression analyses as conducted on Black-owned firms, and discussed at length above. Again, the study looks at two pools of Hispanic-owned prime contracting firms, those HBEs that bid at least once for a County capital construction contract under SIC 15, 16, and 17 during the relevant time period (bidders), and those HBEs that were awarded at least one County capital construction contract over *1568 the relevant time period (awardees). Defs' Prop. Findings 52: Defs' Ex. M-1.
When the HBE bidder and awardee data is disaggregated by type of construction contract the results are as follows:
Fiscal Years 1989 1991
HBE Bidders(%) HBE Awardees (%) Contract $(%)
SIC 15 31.0 33.0 15.0
SIC 16 23.2 21.9 14.2
SIC 17 28.6 31.1 7.2
Defs' Ex. M-1.
For SIC 15, the disparity between HBE bidders (31.0) and contract dollars awarded to HBEs (15.0) is 48.4, which is statistically significant at a level of 2.8 standard deviations. The disparity between HBE awardees (33.0) and contract dollars (15.0) is 45.5, significant at 2.9 standard deviations. The disparity between HBE bidders in SIC 16 (23.2) and contract dollars awarded in SIC 16 (14.2) is 61.2, at 2.1 standard deviations of statistical significance. The disparity for awardees in SIC 16 (21.9) versus contract dollars awarded in SIC 16 (14.2) is 64.9 at 1.8 standard deviations. Finally, for SIC 17, the disparity between HBE bidders (28.6) and contract dollars awarded (7.2) is 25.3, with a standard deviation of 3.5, and for HBE awardees (31.1) versus contract dollars (7.2) is 23.2, at 3.6 standard deviations.
The regression analyses run on these disparity indices included the same two explanatory variables to account for firm size: total awarded value of all contracts bid on; and largest single contract awarded. Defs' Ex. M-2. When controlling for firm size by regressing for total awarded value of all contracts for the years 1989 1991, the only SIC category in which the award disparity appears to be statistically significant is SIC 17, at 3 standard deviations. Id. For SIC 15, the regression results in a standard deviation of 1.1. Id. For SIC 16, the regression indicates that there are no statistically significant disparities in contract dollars awarded. Id.
When controlling for firm size by regressing for largest contract awarded in the years 1989 1991, the disparities are statistically significant for SIC 15, at 2.8 standard deviations, and SIC 17, at 2.4 standard deviations. Defs' Ex. M-2. The regression analyses for SIC 16 suggests that HBEs are awarded 20% more contract dollars, significant at 1 standard deviation. Id.
In 1993, when disaggregated, the HBE County contracting awards trend was:
HBE Bidders(%) HBE Awardees(%) Contract $(%)
SIC 15 31.7 33.9 24.4
SIC 16 22.5 26.5 18.2
SIC 17 29.5 30.0 32.7
Defs' Tr. Ex. M-3.
For SIC 15, the disparity between HBE bidders (31.7) and contract dollars awarded to HBEs in SIC 15 (24.4) is 76.9, which is statistically significant at a level of 1 standard deviation. The disparity between HBE awardees in SIC 15 (33.9) and contract dollars (24.4) is 72.1, significant at 1.1 standard deviations. The disparity between HBE bidders (22.5) and contract dollars awarded (18.2) in SIC 16 is 80.8, which has no statistical significance, and for awardees in SIC 16 (26.5) versus contract dollars (18.2) is 68.8 at 1.3 standard deviations. Finally, for SIC 17, the disparity between HBE bidders (29.5) and contract dollars awarded (32.7) is 110.9, with no statistical significance, and for HBE awardees in SIC 17 (30.0) versus contract *1569 dollars (32.7) is 109.1, with no statistical significance.
The regression results for the disparity indices calculated for the year 1993 are in no case statistically significant at a level of two standard deviations or more. Defs' Tr. Ex. M-4. In fact, in SIC 15, for both total awarded value of all contracts and largest contract awarded, disparities are positive with statistical significance of more than 1.5 standard deviations. Id.
Comparing the proportion of HBE bidders to the proportion of HBEs awarded contracts in both time periods yields similar results.[30] For SIC 15 HBEs received more than their proportional share of contracts, with the disparity indices at approximately 1.06-1.07. For SIC 16 the disparity values are slightly below parity, between 94-96. Lastly, for SIC 17 HBEs receive a proportionate share of the contracts they bid on, with disparity indices between 1.01-1.08. Thus, Hispanic-owned firms win more than their proportional share of contracts in nearly every SIC category for the relevant time periods.
The 1989-1991 disparity analysis is ambiguous concerning the extent of disparities between the proportion of HBEs awarded County contracts and the proportion of County contract dollars that go to HBEs. Defs' Ex. M-2. SIC 17 is the only category that shows statistical significance when both total awarded contracts and largest awarded contract are analyzed. The regression for SIC 16 shows no statistically significant disparity when either size variable is analyzed. SIC 15 is significant to a level of 1.8 standard deviations when total awarded contracts is regressed, and is significant at 2.9 standard deviations when largest contract awarded is regressed.
The 1993 regression analyses do not support a prima facie of discrimination. Defs' Ex. M-4. None of the SIC categories show a statistically significant disparity of two standard deviations or greater. In fact, HBEs in SIC 15 are receiving proportionally more contract dollars than their percentage in the population of total firms, both when regressed for total contracts awarded and largest contract awarded.
To the extent that there is evidence that HBEs get smaller than average dollar awards over the years 1989-1991, this is no longer the case in 1993. The fact that HBEs tend to be smaller, newer firms could be driving the 1989-1991 results. The 1987 Dade County construction business census and survey data show that HBEs have, on average, 4.3 employees, a payroll of $70,893 and sales of $427,032. Comparatively, non-MWBEs have 14.1 employees, a payroll of $272,839, and sales of $1,268,291. Examination of the size of HBEs based upon the dollar value of contracts bid upon by HBEs, demonstrates this as well. See, Defs' Ex. A-1, Tables C-42, at 3-4. In SIC 15, between 1989 and 1992, out of 45 total HBE bids, 22 were for contracts under $500,000, 17 were for contracts between $500,000 and $2,999,999 and 6 were for contracts over $3 million. In SIC 16, between 1989 and 1992, out of 36 total HBE bids, 14 were for contracts under $500,000, 15 were for contracts between $500,000 and $2,999,999, and 7 were for contracts over $3 million. In SIC 17, between 1989 and 1992, out of 30 total HBE bids, 25 were for contracts under $500,000, 3 were for contracts between $500,000 and $2,999,999, and 2 were for contracts over $3 million. These data show that more than 50% of HBE bids are in the lowest dollar value category, contracts under $500,000.
Overall, defendants' data demonstrate support for the conclusion that HBEs are smaller, newer firms, suffering the same obstacles all small and new firms suffer, not that they are being discriminated against due to their ethnicity.
The disparity and regression analyses for Hispanic subcontractors was conducted using the same data and methodology as that used for the BBEs described above. Only SIC 17 displayed a disparity at a level of statistical significance above 2, at 2.2 standard deviations. *1570 Defs' Ex. M-5. The regression analysis conducted to account for size again indicated no disparities of statistical significance for SIC 15 and 16. Defs' Ex. M-6. For SIC 17, the regression resulted in a statistical significance of just under 2 standard deviations.
Of course, the same issues of validity and accuracy of the underlying data that were discussed under the BBE subcontractors analysis are present here. First, the universe identified as subcontractors may be overstated. The firms evaluated in the MRD study had filed a release of lien with the County on at least one occasion. The study's aggregate numbers would include those prime contractors that acted as subcontractors on their own projects, or on the projects of other prime contractors. Second, the sales and receipts variable relied upon by the defendants as an indicator of disparity, and therefore discrimination, included income from all sources derived from anywhere in the United States. Thus, the analysis itself can shed little light upon subcontracting trends in Dade County, because the results of subcontracting choices in other jurisdictions may skew the analysis of what is occurring in Dade County.
The marketplace analysis conducted on HBEs also indicates that, on average, HBEs tend to have lower sales and receipts than non-MWBEs. Defs' Ex. M-7 ("W" files). Again, given that non-MWBEs tend to be larger than HBEs, this is not unexpected. After regressing for number of employees, a proxy for firm size, the disparities in sales and receipts were statistically significant for both SIC 15 and SIC 17. Defs' Ex. M-8. Because of the general nature of the data used in this marketplace analysis, and their inapplicability to the jurisdictional inquiry this Court is confined to, the Court does not find this analysis demonstrative of discrimination in Dade County.
The MRD study also analyzes female-owned firms to assess whether disparities exist in the award of contract dollars and conducts regression analyses to determine whether the disparities can be attributed to discrimination. The study looks at two pools of female-owned prime contracting firms, those WBEs that bid at least once for a County capital construction contract under SIC 15, 16, and 17 during the relevant time period (bidders), and those WBEs that were awarded at least one County capital construction contract over the relevant time period (awardees). Defs' Prop. Findings 61: Defs' Ex. N-1.
When the WBE bidder and awardee data is disaggregated by type of construction contract the results are as follows:
Fiscal Years 1989 1991
WBE Bidders(%) WBE Awardees(%) Contract $(%)
SIC 15 6.9 6.0 1.0
SIC 16 3.2 2.2 2.9
SIC 17 13.3 13.5 4.4
Defs' Ex. N-1.
For SIC 15, the disparity between WBE bidders (6.9) and contract dollars awarded to WBEs (1.0) is 14.6, which is statistically significant at a level of 2.1 standard deviations. The disparity between WBE awardees (6.0) and contract dollars (1.0) is 16.8, significant at 1.9 standard deviations. There is no statistically significant disparity for WBEs in SIC 16. For SIC 17, the disparity between WBE bidders (13.3) and contract dollars awarded (4.4) is 33.2, with a standard deviation of 1.9, and for WBE awardees (13.5) versus contract dollars (4.4) is 32.7, at 1.9 standard deviations.
The regression analyses run on these disparity indices included the same two explanatory *1571 variables to account for firm size: total awarded value of all contracts bid on; and largest single contract awarded. Defs' Tr. Ex. N-2. When controlling for firm size by regressing for total awarded value of all contracts for the years 1989-1991, the disparities for SIC 15 exhibit a statistical significance of 1.1, and for SIC 16 display no statistical significance at all. The regression for SIC 17 indicates that WBEs receive 3 times more contract dollars than their proportion in the relevant population, and this calculation is statistically significant at 2.2 standard deviations. When controlling for firm size by regressing for largest contract awarded in the years 1989-1991, none of the SIC categories indicate any statistical significance whatsoever. Defs' Tr. Ex. N-2.
In 1993, when disaggregated, the WBE County contracting awards trend was:
WBE Bidders(%) WBE Awardees(%) Contract $(%)
SIC 15 13.5 6.1 0.9
SIC 16 9.2 5.7 5.3
SIC 17 9.8 15.0 25.4
Defs' Ex. N-3.
For SIC 15, the disparity between WBE bidders (13.5) and contract dollars awarded to WBEs in SIC 15 (0.9) is 6.3, which is statistically significant at a level of 2.8 standard deviations. The disparity between WBE awardees in SIC 15 (6.1) and contract dollars (0.9) is 13.8, significant at 1.6 standard deviations. The disparity between WBE bidders (9.2) and contract dollars awarded (5.3) in SIC 16 is 57.3, which is significant to a level of 1 standard deviation. For awardees in SIC 16 (5.7) versus contract dollars (5.3) is 91.4 and not statistically significant. Both WBE bidders and awardees in SIC 17 receive twice as many contract dollars than would accord with their numerical proportion.
The regression results for the disparity indices calculated for the year 1993 are statistically significant for SIC 15 bidders and awardees to a level of 3.3 and 3.1 standard deviations, respectively. None of the other SIC categories exhibit any statistical significance when regressed for either total awarded value or largest contract awarded.
The disparity analysis comparing the number of WBE bidders and the number of WBEs that received contracts awards presents a mixed picture.[31] For the time period of 1989 through 1991, in SIC 15 the disparity index is 87, in SIC 16 is 68 and in SIC 17 is at parity 1.01. In 1993 the disparity index for SIC 15 had dropped to 45, for SIC had remained nearly constant at 62, and for SIC 17 had jumped to 1.53.
The 1989-1991 disparity analyses demonstrate no statistically significant disparities between the proportion of WBEs awarded County contracts and the proportion of County contract dollars that go to WBEs. SIC 15 is the only category that shows statistical significance when both total awarded contracts and largest awarded contract are analyzed for 1993. The regressions for SIC 16 and 17 show no statistically significant disparities when either size variable is analyzed for 1993.
Again, size and experience is likely a determining factor in explaining why WBEs get smaller than average dollar awards and win fewer contracts in SIC 15 (this being the only SIC category that is consistent in demonstrating disparities). The 1987 Dade County construction business census and survey data show that WBEs have, on average, 6.6 employees, a payroll of $113,761 and sales of $632,500. For comparison purposes, non-MWBEs firms numbers are more than double in each category; employees, payroll, and sales. An evaluation of the size contracts WBEs tend to bid on strongly suggests WBEs tend to be smaller firms. See, Defs' Ex. A-1, Tables C-42, 3-8. In SIC 15, between *1572 1989 and 1992, out of 3 total WBE bids, 2 were for contracts under $500,000, 1 was for a contract between $500,000 and $2,999,999, and no WBEs bid on contracts over $3 million. In SIC 16, between 1989 and 1992, out of the 4 total WBE bids, 2 were for contracts under $500,000, 0 were for contracts between $500,000 and $2,999,999, and 2 were for contracts over $3 million. In SIC 17, between 1989 and 1992, out of the 5 total WBE bids, all 5 were for contracts under $500,000. Thus, 75% of all WBEs bids between 1989 and 1992 were under $500,000.
Based on the totality of the data, the Court does not find that the MRD prime contracting data provide evidence sufficiently probative to support the County's stated rationale that discrimination is negatively impacting female-owned construction firms in Dade County.
The disparity and regression analyses for female subcontractors were conducted using the same data and methodology as that used for the BBEs and HBEs. Only SIC 15 indicated a disparity statistical significance, at 2.5 standard deviations. Defs' Ex. N-5 ("X" file). The regression analysis conducted to account for size displayed no disparities of statistical significance for any SIC category. Defs' Ex. N-6. Given this Court's concerns about the appropriateness of the underlying data in depicting subcontracting trends in Dade County, and the lack of statistically significance after regressing for size, this analysis does not provide an adequate basis for the County's stated rationale of discrimination in County contracting towards WBEs.
The marketplace analysis conducted on WBEs suggests that the only statistically significant disparity index for sales and receipts shows up in SIC 15. Defs' Ex. N-7 ("W" files). The regression analyses demonstrate no statistically significant disparities for any SIC category. Defs' Ex. N-8. Again, this portion of the MRD study does not aid in substantiating defendants' assertion that female-owned firms are the victims of discrimination in Dade County.
The defendants introduced a statistical analysis prepared by Mr. Jon Wainwright[32] (hereinafter Wainwright study). The Wainwright study seeks to evaluate rates of business ownership in the construction industry for women, Blacks and Hispanics as compared to that of white males in Dade County (business ownership analysis). Tr. 35; Defs' Tr. Ex. F. It also analyzes the disparities in income between MWBEs and non-MWBEs in the local construction industry (business owner income analysis). Id. The data Wainwright uses come from the 1990 Public Use Microdata Sample (PUMS), which is derived from the decennial census. Tr. 33-34.; Defs' Tr. Ex. F at 2. The PUMS data reflect a sample of individuals located in Dade County who identify themselves as self-employed in the construction industry. Tr. 37.
The Wainwright study first identifies disparities in both the rate of business ownership and the income levels of MWBEs versus non-MWBEs, then performs regression analyses on these disparities to determine the extent to which factors other than race, ethnicity, and gender could account for the identified disparities. The regression analyses included explanatory variables known as "human capital" variables and "financial capital" variables. Tr. 38. Human capital variables include personal characteristics such as years of education, years of labor market experience, marital status, etc. Defs' Tr. Ex. F, Tables 1 and 2.[33] Financial capital variables include personal financial attributes such as interest and dividend income, residual household income after personal income, whether the business is incorporated, whether the individual owns his own home free and clear, etc. Id.
*1573 The business ownership analysis evaluates whether Blacks, Hispanics, and women in Dade County own relatively fewer businesses than similarly situated non-Hispanic white males. Tr. 39. Similarly situated means individuals with similar human and financial capital attributes. The analysis indicates that white males go into the construction industry in Dade County at a rate of 21.8%, while Blacks form construction businesses at a rate of 10.2%, for Hispanics the rate is 25%, and for women it is 13.2%. Defs' Tr. Ex. F, Table 3. Using regression analysis, Wainwright estimated that Blacks with approximately the same human and financial characteristics as white males would be expected to enter the construction business at a rate of 20.4%, if they were equally rewarded for these human and financial attributes. This results in a disparity of 50.3, which is statistically significant at a level of 3 standard deviations. Under the same analysis, Hispanics would enter the construction business at a rate of 30.2%, a disparity of 82.6 and significant at 3 standard deviations. Women, if rewarded on par with white males, would enter the construction business at a rate of 19.8%, resulting in a disparity of 66.8, which is also statistically significant at 3 standard deviations.
The business owner income analysis examines whether, once MWBEs are in the construction business in Dade County, gender, race or ethnicity impact the success of the business, as measured by personal income of the owner. Tr. 47; Defs' Tr. Ex. F. The regression analysis demonstrates that financial and human capital characteristics substantially impact the success of a construction business as measured by the owner's personal income. After regressing for various human and financial capital variables, the analysis indicates that for WBEs, the estimated income disparity index is 96.4, with no statistical significance. For Hispanics, the income disparity is 81.2 at a level of 2 standard deviations. For BBEs the income disparity is 72.2, again with a statistical significance of 2 standard deviations.
Mr. Wainwright also testified regarding the results of a study he participated in under the direction of Ph.D. economists Drs. Brimmer and Marshall. (hereinafter the Brimmer study) Tr. 67; Defs' Ex. E. This analysis focused solely on BBEs. The Brimmer study was conducted for Dade County in 1991 using SMOBE data, the Survey of Minority and Women Owned Businesses, conducted by the U.S. Census, from the years 1977, 1982, and 1987. The Brimmer study compares the proportion of BBEs in the Dade County construction industry and the proportion of sales and receipts received by BBEs in Dade County with the total number of construction firms located in Dade County and the total sales and receipts earned by these firms. Examining the disparity indices calculated for SIC 15, 16, and 17 indicate that for BBEs the disparities in sales and receipts were substantial for the years 1977 and 1987, never reaching more than 58. Defs' Tr. Ex. E. Anomalously, for the year 1982, BBE sales and receipts were at or substantially above parity for SIC 15, 16 and 17. Id.
The major drawback of this analysis is that the SMOBE data relied upon do not include information such as firm size, number of employees, etc., thus the Brimmer study does not contain regression analyses to control for neutral variables that could account for these disparities. Mr. Wainwright himself noted the "existence of a disparity doesn't mean necessarily that it's related to discrimination ... our hypothesis is that business ownership is affected by not only discrimination, but perhaps a wide variety of other possible factors as well." Tr. 34, 40.
The Court finds neither the Wainwright study nor the Brimmer study particularly probative on the central issue presented by this case: the extent to which Dade County is either actively or passively participating in discrimination which is occurring in the local Dade County construction industry. The census data used in both studies simply represent individuals or firms located in Dade County which list themselves as being in the business of construction. The census data do not identify whether these entities have ever done work specifically for the County, or to what degree their reported sales or income stems from private sources versus public sources, much less whether the earnings are primarily a result of work done for Dade *1574 County versus Broward County, Palm Beach County, or some other Florida locale, or even sites outside Florida. This lack of specificity makes it difficult, if not impossible, to draw accurate conclusions concerning whether Dade County is itself a participant in gender, racial, or ethnic discrimination to the extent that justifies its use of race, ethnicity, and gender-conscious remedial programs.
With regard to Wainwright's business formation analysis, while this analysis may indicate that a general, societal type of discrimination is possibly affecting the rate that minorities and women tend to form businesses in the construction industry, it is not the type of particularized evidence that is required to provide a strong basis in evidence for the County's race and ethnicity-conscious contract award process, which is aimed at MWBEs which are already in business and qualified to perform the work. In addition, there is substantial evidence that minorities and women have been entering the construction business in substantial numbers over the past decade, which undermines the assertion that significant barriers currently exist to MWBEs in entering the construction industry. See, e.g., Plntfs' Ex. 68.
Plaintiffs point out that the Brimmer study displays an interesting pattern over the time period for which census data are available. In 1977, according to the study, BBEs account for disproportionately fewer sales and receipts than would be expected given their proportion of contractors county-wide. In 1982, BBEs account for a disproportionately greater share of sales than their numbers would predict. Then, even though BBE sales increased in every SIC category between 1982 and 1987, by 1987 BBEs were again receiving disproportionately fewer sales and receipts in proportion to their percentage of contractors in Dade County. This seems to be due to the fact that, as Mr. Wainwright acknowledged, the number of BBEs grew faster than did their share of sales and receipts between 1982 and 1987. Plaintiffs cite to this phenomenon as support for their case in two ways. First, they contend that this robust growth in BBEs is evidence of the favorable work environment in the construction industry towards MWBEs, not evidence of a discriminatory atmosphere. Second, plaintiffs assert that this proves that BBEs are indeed newer, smaller firms able to compete effectively for smaller, rather than larger, contract awards.
Plaintiffs also take issue with the human and financial capital variables Wainwright uses in his regression analyses. For example, Wainwright obtained his measure for an individual's work experience by subtracting the individual's years of schooling from his age. This approach does not capture the individual's actual experience in the construction industry. For instance, two individual's who have been out of school for twenty years will be designated as having the same amount of experience. This could result in misleading comparisons if one individual had worked in a field entirely unrelated to the construction industry for ten of those twenty years while the other had dedicated the entire twenty years to the business of construction.
The Court is troubled by Mr. Wainwright's failure to include age in his litany of human capital variables. See, Defs' Tr. Ex. F, Tables 1 and 2. Mr. Wainwright himself said that an individual's age influences self-employment rates. Tr. 110. He also noted that, generally, Blacks are younger as a population than Whites. The median age for Blacks "is lower by several years." Id. "Therefore, other things being equal ... we would find a disparity driven strictly by age, because age does matter." Id. Hispanics may well share this younger median age characteristic with Blacks. Experience would not seem to be an accurate proxy for age because this number is obtained, as discussed above, by subtracting years of schooling from age. Thus, two individuals with ten years of experience are not necessarily the same age. One could be 28, having completed high school at the age of 18 and started work immediately. The other could be 34, having completed high school, college, and graduate school before beginning his work experience.
Wainwright included a variable to account for an individual's ability to speak English as a personal characteristic that impacts business *1575 formation rates. In his regression analysis this variable turns out to be statistically significant in predicting self-employment rates but it has a negative sign. Thus, individuals who speak English poorly are more likely to form a construction business. This seems counter-intuitive. Plaintiffs take issue with this notion and note that this assumption has the effect of inflating the number of Hispanics predicted by Wainwright to go into the construction business. Plntfs' Ex. 55 at 23. This is the case because significantly more Hispanics speak English poorly than do either Blacks or Whites. Plaintiffs' expert, Dr. John Lunn, reran the regression analysis dropping the English variable and found a significant decrease in the number of Hispanics predicted to form their own construction business, resulting in a positive disparity index of 1.21 for Hispanic males. Clearly, the choice of explanatory variables can effect the outcome of regression analyses.
Plaintiffs' expert Dr. La Noue makes the observation that the Wainwright study did not attempt to categorize business owners by the SIC category in which they do business. Plntfs' Ex. 54 at 69. This is germane to the income disparity analysis because comparing the income of a business owner who is in the heavy construction field (SIC 16) with the income of a business owner who is in the painting and wallpapering business (SIC 17) is akin to comparing apples with oranges. Plaintiffs' introduced evidence indicating that women and minorities are more heavily concentrated in certain fields, such as painting, wallpapering, and decorating, than in others, such as heavy construction. See, Plntfs' Ex. 65. Thus, this criticism may be an important one in terms of explaining some of Wainwright's disparities.
It is clear that the County must make "particularized" findings of discrimination in the local construction industry in order to carry its burden of proof in supporting the need for its affirmative action programs. See Croson, 488 U.S. at 500, 109 S. Ct. at 725 (legislative body may not rest upon generalized assertions); see also H.K. Porter Co. v. Metropolitan Dade County, 975 F.2d 762, 766 (11th Cir.1992); Cone Corp., 908 F.2d at 913. Here, the Court finds the County's statistical evidence to be in conflict. Analyzing the disparity numbers alone, it is apparent that MWBEs are, for the most part, being awarded their proportionate share of Dade County contracts. Presumably because this comparison provides no support for the use of race and ethnicity-conscious measures in the Dade County contracting process, defendants have instead relied upon a comparison of the proportion of MWBEs with the proportion of County contract dollars they receive. Without further elucidation, this disparity calculation displays significant statistical disparities in the proportion of contract dollars being awarded MWBEs for certain SIC categories. Notably, both parties' experts agree that the presence of disparities does not signify the presence of discrimination.[34] Both sides recognize the use of regression analyses to account for the non-discriminatory factors that could affect the distribution of contract dollars in the construction industry, and to illuminate the accuracy of the disparity figures.
The regression analyses are also conflicting, and do not provide a strong basis in evidence to support the County's affirmative action program in its present form. For example, the regression analysis for BBEs shows statistically significant disparities in contract dollars awarded in SIC 15 between the years 1989 and 1991. The significant disparities have disappeared by 1993. Thus, it is highly questionable whether race-conscious measures are appropriate for BBEs performing work in SIC 15 at this point in time, and undeniably doubtful that race and ethnicity-conscious measures should be applied to BBEs operating in any other SIC category. Regression analyses for HBEs show no statistically significant dollar award disparities in 1993, and inconsistently significant disparities for SIC 15 and 17 between 1989 and 1991. WBEs show statistically significant disparities in dollar awards in SIC 15 for 1993, while at the same time receiving nearly twice their proportionate share of contract *1576 dollars in SIC 17. The only statistically significant disparity reported for WBEs from 1989 to 1991 was again a positive disparity in SIC 17, with WBEs being awarded approximately three times their proportionate share in contract dollars. Even analyzed under the intermediate scrutiny standard, these numbers do not, in the Court's opinion, make a case for gender-conscious measures.
The Third, Ninth and Tenth Circuits have discussed the manner in which a municipality's statistics may be rebutted. "First, rebuttal evidence may consist of a neutral explanation for the statistical disparities." Coral Const. Co. v. King County, 941 F.2d 910, 921 (9th Cir.1991). Second, "the rebutting party may wish to attack the statistics themselves. This may be accomplished by (1) showing that the statistics are flawed; (2) demonstrating that the disparities shown by the statistics are not actionable or significant; or (3) presenting contrasting statistical data." Id. Accord, Contractors Ass'n v. Philadelphia, 6 F.3d at 1007; Concrete Works, 36 F.3d at 1530.
Plaintiffs have succeeded in advancing and supporting a race-neutral explanation for the apparent statistical disparities: the disparate sizes between MWBEs and non-MWBEs. Plaintiffs have supported this explanation by providing evidence that non-MWBEs tend to be more experienced firms, on average having been in existence longer, having more employees, larger sales and receipts and greater net worth. The evidence also substantiates the fact that nonprocurement contracts are large, the average being almost $3 million. The combination of these two factors provides a plausible rationale for the trend identified by defendants' statistics, that non-MWBEs tend to receive a disproportionately greater share of the contract dollars awarded by Dade County.
Plaintiffs have also attacked the statistical analyses and underlying data which defendants claim demonstrate disparities and consequently discrimination. It is beyond dispute that defendants' disparity indices alone do not provide evidence of discrimination. Only after these calculations are subjected to the rigor of regression analyses do they have the potential to provide a strong basis in evidence of the existence of discrimination. As has been shown above, after application of regression in order to control for firm size and capacity, most of defendants' calculated disparities lose statistical significance, and therefore are not actionable.
The subcontractor statistical analysis has been demonstrated to be flawed, due to the infirmities of the underlying data. The pool identified as subcontractors are identified as such because they have filed at least one release of lien with the County. It is very likely that many general contractors have filed at least one release of lien with the County, which seriously weakens the basis for this "subcontractors" study. More damaging to the force of this study is that the sales and receipts variable, which is the basis for the comparison that underlies the disparity numbers, is not limited to either funds derived from subcontracting work or to the jurisdiction at issue, Dade County. The limitations of the underlying data render the defendants' subcontractor study ineffective as an aid in demonstrating the occurrence of discrimination in the local construction industry.
Plaintiffs have demonstrated that the Wainwright analysis suffers from similar disabilities due to the underlying data. Indeed, Mr. Wainwright himself acknowledged that the role of his study would be an adjunct to the disparity and regression analyses that must underpin a state or local affirmative action program. An affirmative action program "is valid only if actual, identifiable discrimination has occurred within the local industry affected by the program." Coral Const., 941 F.2d at 916; see also Croson, 488 U.S. at 505, 109 S. Ct. at 728 ("[N]one of the evidence presented by the city points to any identified discrimination in the Richmond construction industry. We, therefore, hold that the city has failed to demonstrate a compelling interest in apportioning public contracting opportunities on the basis of race."). Statistical analyses purporting to show general trends or marketplace phenomena do not suffice to evince "actual, identifiable discrimination" in the "local industry affected."
*1577 The Eleventh Circuit has emphasized the "fact-intensive" nature of this inquiry, and has cautioned that statistical analyses "should not be accorded talismanic significance." Peightal II, 26 F.3d at 1554, 1556. See also, Coral Const. Co. v. King County, 941 F.2d 910, 919 (9th Cir.1991) ("[S]tatistics are not irrefutable; they come in infinite variety and, like any other kind of evidence, they may be rebutted. In short, their usefulness depends on all of the surrounding facts and circumstances.") (quoting International Brotherhood of Teamsters v. United States, 431 U.S. 324, 340, 97 S. Ct. 1843, 1856-57, 52 L. Ed. 2d 396 (1977).) "Ultimately, whether a strong basis in evidence of past or present discrimination exists, thereby establishing a compelling interest for the municipality to enact a race-conscious ordinance, is a question of law." Concrete Works, 36 F.3d at 1522. "Underlying that legal conclusion, however, are factual determinations about the accuracy and validity of a municipality's evidentiary support for its program." Concrete Works, 36 F.3d at 1522. This Court has determined that the accuracy and the validity of the County's statistical evidence fails to form a strong basis in evidence for the conclusion that discrimination has occurred in the Dade County construction industry.
Defendants and intervenors introduced a myriad of anecdotal evidence, both in the form of testimony and affidavits, to support their case. The County presented testimony from two County employees. Herbert Johnson, who has worked for the County in various staff and managerial positions in the area of capital construction for 15 years, is currently the Director of the Dade County Performing Arts Center construction project a $170 dollar County construction project. Tr. 294-95. Mr. Johnson also serves as County chairperson of the MWBE program review committee. Tr. 296.
In his testimony, Mr. Johnson made several key points. He stated that the contracting process in Dade County is decentralized and complex, with the result that County employees involved in the contract award process have substantial discretion in decision making. Tr. 313. He also testified that in his view, it is difficult for new construction firms to break into the County contracting process, because the system is biased towards the status quo, i.e. firms with a proven track record. Tr. 314. The witness also stated that he believed that any discrimination that was occurring within the County was unintentional. Tr. 329. Mr. Johnson testified an incident that he described as an instance of discrimination in the private sector. He recalled that, five or six years ago, a minority subcontractor was given a price by an equipment supplier that was substantially higher than the price given by the same supplier to a non-minority subcontractor for the same piece of equipment. Mr. Johnson could not recall the name of either subcontractor involved in this incident, of the supplier, or even whether the minority subcontractor was Black or Hispanic. Mr. Johnson was reluctant to ascribe this incident to discrimination, instead he attributed it to the nature of individuals' relationships with one another. Tr. 324. Mr. Johnson acknowledged that the County did not investigate the incident, took no disciplinary action against the supplier, and he admitted that he did not know if the subcontractors' credit history or volume of purchases impacted the price quoted by the equipment supplier. Tr. 349.
Next the County presented testimony from Mr. Gregory Owens, a County employee who was the director of the County's Department of Business and Economic Development (DBED) program, which implements the County's MWBE programs, between 1991 and 1995. Tr. 362. Mr. Owens also testified to the fact that, in his opinion, the great amount of discretion County employees have within the County construction contracting process allows discrimination which he believes occurs in that process. Tr. 364-65. Mr. Owens testified that in the past he has heard County employees make comments concerning the inability of Black, Hispanic or women-owned firms to meet tight time lines, or that their prices are too high, or that they simply aren't able to perform the work in question. Tr. 366. He did not identify which employees made these statements but Mr. Owens testified that he did not believe *1578 these employees had ever been reprimanded for these statements. Another example of discrimination in the County contracting process, according to Mr. Owens, occurred when DBED designated a contract for a BBE subcontractor participation goal, only to find out that when the department in charge of that particular contract actually advertised it for bid, the contract did not include the BBE goal. Mr. Owens admitted that a small or new non-minority male firm would likely experience the same type of treatment described above, although, in his opinion, non-MWBEs would be able to overcome these obstacles more quickly than would MWBEs, because the system is biased in favor of non-MWBEs. Tr. 373. Finally Mr. Owens described a number of other occurrences that he implied could be a result of discrimination, such as slow payment by the County to MWBEs, or by prime contractors to MWBE subcontractors. Mr. Owens could not testify to whether this was happening to non-MWBEs with any frequency. He also testified to the fact that MWBEs often complain that they get lengthy punch lists lists by the prime contractor or project supervisor of items that were not performed to specification and must be redone when the non-MWBEs on the same project were not getting the punch lists of that length. Both County witness discussed the difficulty MWBEs have in getting bonding and financing.
The main points the Court draws from the County witnesses is that the process by which County construction contracts are awarded is fraught with opportunities for individuals to exercise discretion in decision making and that the process is biased against small and new entrants. The defendants ask the Court to make the leap to conclude that because there is individual discretion, there is individual discrimination. Neither County employee could identify instances in which County personnel engaged in discriminatory behavior that lead to a willing, qualified and able MWBE being deprived of a contract award. Rather, they testified in general terms about the fact that County personnel "could" be "predisposed" to view minorities or female contractors less favorably than their white male counterparts, or that County personnel "could" hold negative stereotypes of MWBEs that "could" influence their decision making with regard to contract awards. See, e.g., Tr. 302. This type of speculation does not form a strong basis in evidence of discrimination occurring in the Dade County construction industry.
The County witness were far more persuasive on the point that new construction firms face substantial barriers to entry into County contracting. Herbert Johnson identified many aspects of the County contracting process that work against new firms being awarded County contracts, including the size and scope of the contract award, the way in which the contract specifications are written and the nature of the requirements, etc. Tr. 301-04. The County could certainly take steps to change these aspects of the system in order to encourage minority and new firm participation.
Even if the Court did accept the hypothesis that because there is discretion there is discrimination, this does not justify the use of race, ethnicity, and gender-conscious measures by the County in awarding construction contracts. From the testimony of the County witness it appears that there is no program in place to deal with County employees, contractors, or suppliers who engage in discriminatory conduct. The County could sanction individuals or companies which discriminate against women and minorities. The County could also implement a training program for its employees and others that do business with the County to sensitize these individuals as to the inappropriateness and illegality of racial and gender stereotyping. The County could attempt to standardize and centralize its contracting program to limit the amount of individual discretion currently afforded by the system.
The defendants and intervenors also introduced, via trial and deposition testimony as well as affidavits, the statements of at least 21 contractors and subcontractors who work in the Dade County construction industry. See, Tr. 413-800; Intervenors' Ex. B-1. These witnesses described numerous incidents in which they believe that they were deprived of work because of their race, ethnicity, *1579 or gender. This includes repeated instances in which the project supervisor or site foreman would not deal directly with minority or female business owner but would instead deal with a non-MWBE employee; situations in which suppliers would refuse to quote the MWBE a price or would escalate the price as compared to that quoted a non-MWBE; times when the MWBE knew itself to be the low bidder on a subcontracting project but was not awarded the project, or instances in which the quote given by the MWBE was "shopped" by the prime in order to solicit lower bids from non-MWBE subcontractors; repeated incidents in which a MWBE subcontractor would receive a bid invitation from a general contractor within a day of being due and with a letter of unavailability that the GC would submit to the County; situations in which the MWBE was hired by a prime, most often pursuant to the County MWBE program, and was replaced by a non-MWBE subcontractor within days of starting work on the project.
Defendants also introduced a study based on anecdotal accounts of discrimination that was prepared for a previous case in which Dade County's affirmative action program was challenged. The County retained Dr. Joe Feagin, Chairman of the Department of Sociology at the University of Florida, to survey BBEs in Dade County. Defs' Tr. Ex. B (hereinafter "Feagin report"). Dr. Feagin interviewed 78 construction firms which had been certified by Dade County as Black-owned businesses. According to Dr. Feagin's account of these interviews, the BBEs surveyed reported encountering the same types of obstacles and experiences that were described in the testimony and affidavits introduced in this case by the defendants and intervenors. For example, BBEs in the Feagin report cited difficulty in securing financing and bonding, slower payment by general contractors, unfair evaluations of their performance based on racial stereotypes, lack of access to information from County personnel on contracting processes, and higher prices on equipment and supplies.
Plaintiffs' respond with several points the Court believes to be valid concerning the reliability of this anecdotal evidence. First, whether discrimination has occurred is often complex and requires a knowledge of the perspectives of both parties involved in an incident as well as knowledge about how comparably placed persons of other races, ethnicities, and genders have been treated. Persons providing anecdotes rarely have such information. Attributing an incident to discrimination when the practice is just aggressive business behavior, barriers faced by all new or small businesses, or bad communication is always a possibility.
Second, social scientists are frequently concerned about the problem of "interviewer bias" or "response bias" in any interviewing or survey situation. When the respondent is made aware of the political purpose of questions or when questions are worded in such a way as to suggest the answers the inquirer wishes to receive, "interviewer bias" can occur. If a sample is not carefully constructed, the persons providing the anecdotes may reflect a "response bias" because the persons most likely to respond are those who feel the most strongly about a problem, even though they may not be representative of the larger group.
Third, individuals who have a vested interest in preserving a benefit or entitlement may be motivated to view events in a manner that justifies the entitlement. Consequently, it is important that both sides are heard and that there are other measures of the accuracy of the claims. Attempts to investigate and verify the anecdotal evidence should be made.
The anecdotal evidence strongly suggests that the construction industry is built upon relationships and past experiences. The evidence indicates that general contractors do business with subcontractors and suppliers with whom they have worked before and those they feel have a proven track record. The results in the inevitable conclusion that it is difficult for a new entrant to break in and gain a foothold within the network of established firms. This is not evidence of a pattern or practice of discrimination against BBEs, HBEs, or WBEs. "Anecdotal evidence is most useful as a supplement to strong statistical evidence" and it "rarely, if ever, can ... show a systemic *1580 pattern of discrimination necessary for adoption of an affirmative action plan." Coral Constr., 941 F.2d at 919; O'Donnell, 963 F.2d at 427. Defendants' anecdotal evidence cannot cure the weaknesses of defendants' statistical evidence; it is not the sort of "identified discrimination" contemplated by Croson. See 488 U.S. at 505, 109 S. Ct. at 728.
"The essence of the narrowly tailored inquiry is the notion that explicit racial preferences ... must be only a last resort option." Hayes v. North State Law Enforcement Officers Ass'n, 10 F.3d 207, 217 (4th Cir.1993) (internal quotations and citation omitted). The Croson Court directed that racially preferential programs be reserved for the "extreme case," where "necessary to break down patterns of deliberate exclusion." 488 U.S. at 509, 109 S.Ct. at 730. Governmental entities must proceed very cautiously when imposing race or ethnicity-conscious measures because, "[c]lassifications based on race carry a danger of stigmatic harm. Unless they are strictly reserved for remedial settings, they may in fact promote notions of racial inferiority and lead to a politics of racial hostility." Croson, 488 U.S. at 493, 109 S. Ct. at 721. "[E]ven in the pursuit of remedial objectives, an explicit policy of assignment by race may serve to stimulate our society's latent race consciousness, suggesting the utility and propriety of basing decisions on a factor that ideally bears no relationship to an individual's worth or needs." United Jewish Orgs. v. Carey, 430 U.S. 144, 173, 97 S. Ct. 996, 1014, 51 L. Ed. 2d 229 (1977) (Brennan, J., concurring in part).
Therefore, "[t]o invoke race-conscious relief, a state or local government must utilize means which are specifically and narrowly tailored to accomplish the remedial purpose." Peightal II, 26 F.3d at 1557. Since Croson, the Eleventh Circuit has set out in plain terms the elements that must be evaluated to determine whether a public employer's affirmative action plan is sufficiently narrowly tailored to pass constitutional muster. The four factors that must be considered are: "(1) the necessity for the relief and the efficacy of alternative remedies; (2) the flexibility and duration of the relief, including availability of waiver provisions; (3) the relationship of numerical goals to the relevant labor market; and (4) the impact of the relief on the rights of innocent third parties." Ensley Branch, N.A.A.C.P., 31 F.3d at 1569 (internal cites and quotes omitted); Peightal II, 26 F.3d at 1557.
The "initial inquiry" for this Court is "to determine whether the local government considered the use of race-neutral means to increase minority participation." Id. "While strict scrutiny does not require exhaustion of every possible ... alternative, it does require serious, good-faith consideration of race-neutral alternatives either prior to or in conjunction with implementation of an affirmative action plan." Ensley Branch, N.A.A.C.P., 31 F.3d at 1571 (internal quotations and citations omitted). The evidence here suggests that serious, good-faith consideration of race and ethnicity-neutral alternatives was not undertaken by the County commission. The legislative findings to the BBE ordinance makes the conclusory statement that "race neutral programs cannot address the above problems and do not focus limited County money, efforts and personnel on the problems caused by racial discrimination." Ordinance 94-96, p. 4. This conclusion was drawn based upon an eight page analysis by consultant Andrew F. Brimmer as to the utility and efficacy of race-neutral alternatives. Defs' Ex. A at 102-110. This "analysis" does not identify any specific race-neutral measures that the County ever has, or potentially could, engage in, nor does it even attempt to assess the strengths and weaknesses or potential for success or failure of possible race and ethnicity-neutral alternative. It simply concludes that race and ethnicity-neutral alternatives would not adequately remedy discrimination in Dade County. Defendants also cite as additional support for the Commission's conclusion a 1982 report of the Civil Rights Commission. Defs' Ex. C. This report evaluated the Small Business Administration's efforts at providing financial assistance to *1581 Black businesses from 1968 to 1980. Because only five percent of the SBA financing went to Black business over this time period, the Commission apparently concluded that race-neutral attempts to aid Black business achieve financing do not work. No assessment was undertaken to determine the causes for this apparent lack of success. Nor were strategies discussed concerning the utility of efforts such as outreach programs targeted at minorities in increasing the level of financing going to the minority community. In any event, the Court is not convinced that the track record of the SBA during the 1970's is an appropriate indicator of the potential success of a program undertaken in the 1990's. The Commission made no findings concerning the use of race, ethnicity, or gender-neutral programs in support of either the HBE or WBE programs.
The strongest indicator that race and ethnicity-neutral alternatives have not been considered in a serious and good-faith manner is presented by the testimony of defendants' witnesses at trial. The two senior Dade County managers gave ample evidence that the barriers faced by MWBEs in Dade County construction contracting are faced by all new and small firms entering the market. Both witnesses described a litany of barriers and potential remedies that are not race, ethnicity, or gender based, therefore do not require a race-conscious remedy. Both Mr. Johnson and Mr. Ownes identified the level of discretionary decisionmaking by County employees, the complexity of contract specifications, the difficulty in attaining bonding and financing, unnecessary bid restrictions, bias towards firms with a proven track record, inefficient payment procedures, and lack of information exchange and availability as significant impediments to minority, small, and new firm participation in County contracting. See, e.g., Tr. 303, 313, 316, 364, 373, 380. The Supreme Court has encouraged race-neutral means of addressing similar issues identified in the Richmond contracting system:
[T]he city has at its disposal a whole array of race-neutral devices to increase the accessibility of city contracting opportunities to small entrepreneurs of all races. Simplification of bidding procedures, relaxation of bonding requirements, and training and financial aid for disadvantaged entrepreneurs of all races would open the public contracting market to all those who have suffered the effects of past societal discrimination or neglect. Many of the formal barriers to new entrants may be the product of bureaucratic inertia more than actual necessity, and may have a disproportionate effect on the opportunities open to new minority firms. Their elimination or modification would have little detrimental effect on the city's interests and would serve to increase the opportunities available to minority businesses without classifying individuals on the basis of race. The city may also act to prohibit discrimination in the provision of credit or bonding by local suppliers and banks. Business as usual should not mean business pursuant to the unthinking exclusion of certain members of our society from its rewards.
Croson, 488 U.S. at 509-10, 109 S. Ct. at 730. The record does not indicate that Dade County has even considered, much less tried, many of these race and ethnicity-neutral alternatives to increasing minority participation in County contracting.[35] The technical and financial aid the County does offer has not been evaluated for its effectiveness. Defs' Prop. Findings at 42-3. As the Ninth Circuit has opined:
many of the problems caused by the relative youth of minority-owned firms can be resolved without resorting to stigmatizing and fractionalizing racial classifications. Likewise, a well-conceived race-neutral alternative ensures that the minority beneficiaries of the program are more likely to be the true victims of discrimination; such a program prevents a race-conscious program that merely acts as a windfall to previously established minority firms.
*1582 Coral Constr., 941 F.2d at 923.[36]
The record additionally suggests that where discriminatory behavior has been identified, whether in isolated County employees or private equipment suppliers, the County has taken no steps to reprimand, penalize or denounce the unacceptable conduct. Nor has the County passed ordinances prohibiting contractors from discriminating in their choice of subcontractors or suppliers in discriminating against customers on account of race, ethnicity, or gender. Jt. P-T Stip., Attachment A at 9. Not only does this deprive County personnel of the legal means with which to assail this behavior, it sends a message to the public that this behavior is not unacceptable.
This Court firmly believes that "preferential programs may only reinforce common stereotypes holding that certain groups are unable to achieve success without special protection." Regents of University of California v. Bakke, 438 U.S. 265, 298, 98 S. Ct. 2733, 2757, 57 L. Ed. 2d 750 (1978). "One color of discrimination has been painted over another in an effort to mask the peeling remnants of prejudice past, leaving a new and equally offensive discoloration rather than a clean canvas. The time has long passed for the [County] to strip away the past and adopt fresh, race-neutral selection procedures." Ensley Branch, N.A.A.C.P., 31 F.3d at 1572-73. Accordingly, the Court finds that the County has failed to demonstrate that its affirmative action plan is "designed to further some goal other than outright racial balancing," Cone Corp., 908 F.2d at 914, and that "race-neutral alternatives have not been pursued diligently enough in this case." Id., 31 F.3d at 1571.
The second inquiry this Court must undertake evaluates whether the degree of remedial preference is tied to the effects of past discrimination. Peightal II, 26 F.3d at 1558 "This inquiry analyzes the flexibility and duration of the remedy as well as the relationship of the numerical goals to the relevant labor market." Id. (citing U.S. v. Paradise, 480 U.S. at 171, 107 S. Ct. at 1066).
The County's affirmative action program appears to have the requisite flexibility required under Croson and its progeny. The County MWBE program is applied on a case by case basis with an individual review of each contract before application of any contract measures. There are waiver and appeal provisions after a contract has been designated for contract measures. Plaintiffs' witness James A. Cummings testified that he had successfully appealed a contract with subcontractor goals under these procedures and had never lost a bid when he was the low bidder due to failure to meet a contract measure. Tr. 676, 733-34.
The duration of the program is limited to the extent that the need for the program is reevaluated once every five years, after the publication of the "Survey of Minority Owned Business Enterprises" (SMOBE) by the Census Bureau. In addition, the program's application to a particular market segment automatically expires when statistical evidence of disparity ceases to exist. The Court does not find this proof of the limited duration of the program. Every expert economist who testified in this case stated unequivocally that the existence of numerical disparities do not lead to the conclusion that discrimination exists. This is because simple disparity indices do not account for the myriad factors that can legitimately result in disparities, such as the availability of MWBEs that are actually qualified to perform the contract requirements, the size of a firm, which will impact the dollar value of contracts which can be successfully bid for, the capacity of a firm to handle multiple County contracts at the same *1583 time, etc. Only when these and other factors which affect the qualifications, ability, and willingness of a firm to compete for County construction work are taken into account through using appropriate data and performing regression analyses can the County accurately determine whether there are actual disparities in the number of willing able and qualified MWBEs and the amount of work they are performing in the local construction industry.
The County's affirmative action program fails to establish an appropriate relationship between the numerical goals prescribed and the relevant labor pool. The County's participation goals are 19% for Hispanics, 15% for Blacks, and 11% for women. The County presented evidence at trial that demonstrated that the number of available MWBEs vary dramatically for HBEs, BBEs and WBEs, and between SIC 15, 16 and 17. For example, the County currently enforces a 15% participation goal for BBEs performing work in SIC 16. The County's analysis indicates that BBEs only comprise between 3.4% and 5.2% of the firms doing business in that SIC category.[37] Defs' Ex. L-1. Clearly, the enforcement of a 15% participation goal in this SIC category does not represent a rational relationship between the numerical goal and the relevant labor pool. Likewise, Hispanics make up more than 22% of each SIC category, and over 30% in SIC 15. Defs' Ex. M-1. Again, the relationship between the County-imposed goal of 19% and the actual prevalence of HBEs in the various SIC categories is irrational. Lastly, WBEs are no more than 3.2% in SIC 16 and only 6.9% in SIC 15, far below the participation goal of 11% imposed by the County. Defs' Ex. N-1. The Court finds that the County's participation goals simply do not have a reasonable basis given the County's own evidence concerning MWBEs participation in the various construction SIC categories.
Finally, the Court must determine whether the impact of the relief upon third parties is unduly burdensome. Peightal II, 26 F.3d at 1561. "[W]hile attempting to eliminate the vestiges of past discrimination against persons of one race, government must minimize the impositions of new discriminations on persons of other races." In re Birmingham Reverse Discrimination Emp. Lit., 20 F.3d 1525, 1544 (11th Cir.1994). "Critical to the analysis of the appropriate burden to be shouldered by non-minorities is the extent to which the relief disrupts settled rights and expectations." Id. Thus, "[d]enial of a future employment opportunity is not as intrusive as loss of an existing job." Wygant, 476 U.S. at 282-83, 106 S. Ct. at 1851. When it comes to winning or losing contracting or subcontracting bid proposals the analysis is not so clear-cut. Testimony at trial suggested that putting together a bid for a subcontracting job of approximately $1 million and can take up to a week to prepare and as much as two to three weeks for a $3-4 million project. Tr. 749, 769. This is a significant time and resource commitment. Because Dade County awards contracts to the lowest bidder, it is fair to assume that firms make a substantial effort to develop the lowest bid possible. The Court seriously questions whether it is an appropriate burden for a non-minority firm to expend this level of effort in developing what turns out to be the low bid, but is then denied the contract award because the contractor was not able to hire the required number of MWBE subcontractors under a participation goal requirement. This entails much more than simply committing half a day to show up at a job interview.
Based on the totality of the considerations demanded by strict scrutiny to ensure that any race-conscious measures employed by state or local governments are narrowly tailored, this Court finds that the Dade County MBE programs are not sufficiently narrowly tailored to overcome plaintiffs' Equal Protection challenge.
In sum, the Court finds that the conflicting statistical evidence presented by the defendants is insufficient to provide the strong basis in evidence necessary to support the use of race and ethnicity-conscious contract measures by Dade County in determining which firms may bid and be awarded County construction contracts. It is likewise insufficient to provide the factual predicate to support the County's state rationale for its gender preference program. "Statistical evidence often does not fully account for the complex factors and motivations guiding employment decisions, many of which may be entirely race [or gender]-neutral." Coral Const., 941 F.2d at 919. The Court believes that is indeed the case. Plaintiffs' have produced both evidence of race-neutral explanations for the disparities reflected in defendants' statistical analyses and evidence of flaws in the data and methodology that underlie defendants' statistics and make the numerical disparities neither significant nor actionable. Id. at 921; Contractors Ass'n v. Philadelphia, 6 F.3d at 1007; Concrete Works, 36 F.3d at 1530.
The proffered anecdotal evidence does not change this result. Without corroboration, the Court cannot distinguish between allegations that in fact represent an objective assessment of the situation, and those that are fraught with heartfelt, but erroneous, interpretations of events and circumstances. The costs associated with the imposition of race, ethnicity, and gender preferences are simply too high to sustain a patently discriminatory program on such weak evidence.
"Without an adequate showing of discrimination, the government's assertion that affirmative action is necessary lacks credibility." Ensley Branch, N.A.A.C.P., 31 F.3d at 1565 (citing Croson, 488 U.S. at 505, 109 S. Ct. at 728). The evidence presented by the defendants does not constitute an adequate showing of discrimination, under either the strict scrutiny or the intermediate scrutiny standards of review. In accordance with the directive of the Eleventh Circuit Court of Appeals, because defendants have "fail[ed] to present strong evidence justifying race-based relief ... the district court must forthwith terminate the race-based affirmative action provisions." Id. at 1568.
THIS CAUSE came before the Court for TRIAL without a jury. The Court has carefully considered the evidence adduced at trial, and has entered separate Findings of Fact and Conclusions of Law. The Court has concluded that defendants have violated, and continue to violate, the Fourteenth Amendment right of plaintiffs to the equal protection of the laws. Accordingly, it is
ORDERED AND ADJUDGED that
1. FINAL JUDGMENT be, and the same is, hereby entered for the plaintiffs and against the defendants.
2. Plaintiffs' members be, and the same are, hereby DECLARED entitled to submit bids to Metropolitan Dade County for construction work and to have those bids considered without regard the racial, ethnic, or gender character of the bidder or of the bidder's subcontractors or suppliers.
3. Metropolitan Dade County's ordinances, resolutions, regulations, and bid specifications which require the use of MBE subcontractors be, and the same are, hereby DECLARED to be unconstitutional in violation of the Equal Protection Clause of the Fourteenth Amendment.
4. The practices, policies, customs, and usages of Metropolitan Dade County that require consideration of racial, ethnic, and gender criteria for determining whether bids submitted for construction work are responsive be, and the same are, hereby DECLARED to be unconstitutional in violation of the Equal Protection Clause of the Fourteenth Amendment.
5. Metropolitan Dade County be, and the same is, hereby PERMANENTLY ENJOINED from using, or requiring the use, of racial, ethnic, or gender criteria in deciding *1585 (i) whether a bid submitted for construction work is responsive, (ii) whether such a bid will be considered, (iii) whether a contract will be awarded to a contractor submitting such a bid.
The Clerk of the Court shall CLOSE THIS CASE and DENY ALL PENDING MOTIONS AS MOOT. The Court reserves jurisdiction to tax costs and fees.
[1] For the remainder of this opinion, the term "Minority Business Enterprise" or "MBE" will be used to refer to the Black and Hispanic Business Enterprise Programs, "WBE" will be used to refer to the Women Business Enterprise Program, and "MWBE" will be used when discussing all three programs at once.
[2] Throughout this opinion, the Court will refer to race and ethnicity as a single concept for purposes of equal protection review. While analytically distinct, both of these characteristics are subject to the same constitutional protection. See Miller v. Johnson, ___ U.S. ___, ___, 115 S. Ct. 2475, 2482, 132 L. Ed. 2d 762 (1995) ("Racial and ethnic distinctions of any sort are inherently suspect and call for the most exacting judicial examination...."), citing Regents of Univ. of California v. Bakke, 438 U.S. 265, 291, 98 S. Ct. 2733, 2748, 57 L. Ed. 2d 750 (1978) (opinion of Powell, J.).
[3] The program permitted the County to waive formal bid procedures and to "set-aside" contracts for bidding solely among Black contractors where, inter alia, "there exists at least three Black prime contractors with the capabilities consistent with the contract requirements." South Fla. Chap., 723 F.2d at 848. The Program also contained a "goals" provision whereby the County could require a certain percentage of a contract's value to be subcontracted to Black contractors. Id., at 846-47.
[4] In Fullilove v. Klutznick, 448 U.S. 448, 100 S. Ct. 2758, 65 L. Ed. 2d 902 (1980), Chief Justice Burger "did not employ `strict scrutiny' or any other traditional standard of equal protection review." City of Richmond v. J.A. Croson Co., 488 U.S. 469, 487, 109 S. Ct. 706, 718, 102 L. Ed. 2d 854 (1989). Since the Court of Appeals decided South Fla. Chap. v. Metro. Dade in 1984, the United States Supreme Court has rejected the standard set out in Fullilove in favor of a strict scrutiny analysis to review affirmative action programs for equal protection violations.
[5] Dade County's MWBE programs are described in defendants' Trial Exhibit U, in Attachment A of defendants' and intervenors' post trial proposed findings of fact and conclusions of law (DE 123), and in Attachment A of the parties' Joint Status Report (DE 32).
[6] An MWBE which has exceeded the size limit may retain its certification for one year and may retain, apparently indefinitely, its certification if it demonstrates that "it continues to experience the kinds of racial [or gender] discrimination addressed by this section." Section 2-8.2(3)(e) of the Code of Metropolitan Dade County (Defs' Ex. U.1).
[7] The 1972 Standard Industrial Classification Manual differentiates within the construction industry between general building construction (SIC 15), heavy construction other than building construction (SIC 16), and specialty trade construction (SIC 17). In addition, Dade County has developed participation goals for at least 20 other classes of contracts, ranging from machinery and computer equipment to amusement and recreation services. Defs' Ex. W.
[8] The County may also waive competitive bidding if there are at least two MWBEs available, neither of which has been awarded a County contract for the goods or services to be advertised in the last three years, and a price analysis has been conducted to assure that the bid is competitive.
[9] In a competitive bidding situation, the contract typically is awarded to the lowest bidder. By reducing the amount of an MWBE bid by up to ten percent for purposes of determining the lowest bid, an MWBE that is not the lowest bidder can win the contract.
[10] Where bid evaluation procedures assign weights to selection criteria, MWBEs are afforded an additional selection criteria of ten percent. Where weights are not assigned, an MWBE selection factor can be the deciding factor in awarding the bid, if all other factors are substantially equal.
[11] The Court wonders, in passing, whether more recent decisions of the Supreme Court have completely closed the door on race-based classifications. In Romer v. Evans, ___ U.S. ___, ___, 116 S. Ct. 1620, 1623, 134 L. Ed. 2d 855 (1996), the Supreme Court appears to adopt Justice Harlan's famed dissent in Plessy v. Ferguson, 163 U.S. 537, 559, 16 S. Ct. 1138, 1146, 41 L. Ed. 256 (1896), stating that "the Constitution `neither knows nor tolerates classes among citizens.'" This same dissent, of course, insists that "[o]ur constitution is color-blind...." 163 U.S. at 559, 16 S.Ct. at 1146. If this phrase truly is law, and not mere precatory rhetoric, no race-based classification can withstand constitutional scrutiny.
[12] The Brimmer study, discussed infra, also indicates that in 1982 BBEs accounted for a disproportionately greater share of sales than their numbers would predict. This fact, obviously disadvantageous to defendants, may be the reason they prefer to focus on post-enactment data.
[13] Several Courts of Appeal have likewise approved the use of disparity indices to demonstrate the underutilization of minorities or women in a particular industry. See, Cone Corp., 908 F.2d at 916 (11th Cir.); Concrete Works, 36 F.3d at 1523 (10th Cir); AGC of California, Inc. v. Coalition for Economic Equity, 950 F.2d 1401, 1414 (9th Cir.1991); Contractors Ass'n, 6 F.3d at 1005 (3rd Cir.); and Stuart v. Roache, 951 F.2d 446, 451 (1st. Cir.1991) cert. denied, 504 U.S. 913, 112 S. Ct. 1948, 118 L. Ed. 2d 553 (1992).
[14] Standard deviation analysis measures the probability that a result is a random deviation from the predicted result the more standard deviations the lower the probability the result is a random one. Social scientists consider a finding of two standard deviations significant, meaning that there is about one chance in 20 that the explanation for the deviation could be random and the deviation must be accounted for by some factor other than chance. A finding of two or three standard deviations (one in 384 chance the result is random) is generally considered highly probative of discriminatory treatment. Peightal II, 26 F.3d at 1556 (citing Waisome v. Port Authority, 948 F.2d 1370, 1376 (2d Cir.1991)).
[15] The "expected value" in a case such as this is that the number of MWBEs willing, able, and qualified to do the work required by the County contracts is proportionate to the number of MWBEs actually being awarded contracts.
[16] In addition, the aggregated data may not be accurate due to a phenomenon known as "Simpson's Paradox." This occurs when categories of disparate entities are grouped together for statistical purposes. When aggregated, the data may indicate that disparities exist even though, when disaggregated, the disparities disappear. The more heterogeneous the groups the less likely the disparity analysis will be reliable. See, Plntfs' Ex. 55 at 3; Tr. 87, 91-92. Here, the Court feels that firms operating in one SIC category are sufficiently different from firms operating in the others that Simpson's Paradox could well affect the aggregated data.
[17] Fiscal year 1992 was not included in the study because the County's normal contracting procedures were suspended during that year in order to perform the unprecedented scale of emergency construction necessitated by Hurricane Andrew, which occurred August 24, 1992.
[18] See, ftnt 10, supra, and accompanying text.
[19] See, ftnt 11, supra.
[20] In SIC 15 BBEs were 17.5% of the bidders and got 24.6% of the contracts awarded, this results in a disparity index of about 140.5. In SIC 16 BBEs were 16.6% of the bidders and received 24.1% of the contracts awarded resulting in a disparity index of approximately 145.2.
[21] BBEs were 13.8% of the bidders and received 15% of the contracts awarded, for a disparity index of approximately 108.7.
[22] For instance, the EEOC's disparate impact guidelines use the 80% test as the threshold for determining a prima facie case of discrimination. 29 C.F.R. § 1607.4(D).
[23] Tr. 202.
[24] Plaintiffs' Exhibit 67, taken from the 1987 Census of Construction Industries, Geographic Area Series and Survey of Minority-Owned and Women-Owned Business Enterprises. The Business Census is conducted once every five years. At the time of trial, the 1992 Business Census had not yet been released. Tr. 66.
[25] In order to determine whether "discrimination" is occurring, the MRD study analyzes whether the fact that a firm is Black- or Hispanic- or female-owned impacts the dollar value of contract awards that firm receives by introducing an explanatory variable for gender and ethnicity into the regression analyses, along with explanatory variables such as size and experience of the contractor. Tr. 144.
[26] Other courts have noted that a level of statistical significance of two standard deviations or greater is compelling, and this Court agrees. See, Peightal II, 26 F.3d at 1556 (Considering "greater than two to three standard deviations" persuasive.); Waisome, 948 F.2d at 1376 ("A finding of two to three standard deviations ... is generally considered highly probative of discrimination.")
[27] SIC 15 showed statistically significant disparities when regressed for both explanatory variables total awarded value of contracts and largest awarded value of contract. Because both of these explanatory variables were proxies for the same phenomenon size of firm the Court does not consider the weight of these two regressions to be additive.
[28] MRD chose to retain subcontracting data from the year 1992 in which Hurricane Andrew struck due to the three year time lag between when a contract is awarded and when the subcontracting work actually occurs. Tr. 287-88.
[29] In Croson, the Court held that nationwide discrimination could not be used to justify Richmond's affirmative action program, and confined its review to the construction industry in the City of Richmond. See 488 U.S. at 499, 109 S. Ct. at 724.
[30] None of the parties calculated these disparities, leaving the Court to undertake this mathematical feat. While disparity calculations are relatively straightforward, regression equations are not and thus were not attempted.
[31] Again, the disparity numbers comparing the proportion of bidders and awardees were calculated by the Court, therefore, regressions have not been conducted on this data.
[32] Mr. Wainwright holds a Masters degree in economics from the University of Texas and is on the verge of completing his doctoral dissertation in the same field. Tr. 31.
[33] On cross-examination, Mr. Wainwright stressed the fact that age is a significant predictor of whether an individual is likely to be self-employed, but age is not accounted for in Wainwright's human capital variables. Tr. 110.
[34] Mr. Wainwright noted that "the simple existence of a disparity doesn't mean necessarily that it's related to discrimination." Tr. 34.
[35] Compare Hillsborough County's MBE program, which, when enacted, included all of the race-neutral measures suggested in Croson. The Eleventh Circuit found this significant in determining the constitutionality of the program in Cone Corp. 908 F.2d at 916.
[36] The ordinance makes no attempt to punish those guilty of any alleged racial discrimination or to reward those who have been victimized by alleged prior discrimination. Clearly, any MWBE that has achieved ownership of a company that bids on county work has established itself beyond the victim stage. The companies whose low bids have been rejected by this program may never have had any role in the alleged prior discrimination. This Court agrees with Justice Scalia in observing that "under our Constitution there can be no such thing as either a creditor or a debtor race." Adarand, ___ U.S. at ___, 115 S.Ct. at 2118 (Opinion of Scalia, J., concurring in part and concurring in the judgment).
[37] 5.2% represents the proportion of BBEs that have bid on a SIC 16 contract, while 3.4% represents those BBEs that have been awarded contracts under SIC 16, which may be a more accurate assessment of the proportion of BBEs that are qualified to perform the work. Defs' Ex. L-1.