Opinion
CIVIL ACTION H-14-1203
03-22-2016
KOSSMAN CONTRACTING CO., INC., Plaintiff, v. CITY OF HOUSTON, Defendant.
ORDER ADOPTING MEMORANDUM & RECOMMENDATION
Pending before the court is the Magistrate Judge's Memorandum & Recommendation (the "M&R") (Dkt. 43), granting defendant City of Houston's ("Houston") motion to exclude (Dkt. 30); recommending that plaintiff Kossman Contracting Co., Inc.'s ("Kossman") motion for summary judgment (Dkt. 22) be granted in part and denied in part; and recommending that Houston's motion for summary judgment (Dkt. 32) be granted in part and denied in part. Kossman filed objections (Dkt. 45), to which Houston responded (Dkt. 46). Having reviewed the M&R, the objections to the M&R, the response, and the applicable law, the court OVERRULES Kossman's objections and ADOPTS the M&R.
I. BACKGROUND
This case involves an equal protection challenge to Houston's Minority and Women Owned Business Enterprise (MWBE) program. Dkt. 1. The MWBE program has been in place in various iterations since 1984 and has been the subject of significant litigation. Id. at 1. The MWBE program requires that a certain percentage of Houston's contracts on construction projects be awarded to businesses owned by women and minorities. Dkt. 22 at 1. Under the MWBE program, a prime contractor must either (1) achieve the MWBE contract goal by hiring sufficient MWBE subcontracting or (2) obtain a waiver of the goal on the ground that it was unable to meet the goal despite its good faith efforts. Dkt. 1 at 5. The current program has been in effect since July of 2013 and sets a 34% MWBE goal for construction projects. Dkt. 22 at 4. Houston set this goal based on a study performed by National Economic Research Associates ("NERA") that was issued on April 18, 2012. See Dkt. 32, Ex. 1.1A. The study analyzed the status of minority-owned and women-owned business enterprises in the geographic and product markets of Houston's construction contracts. Id. The study relied on data acquired from Dun & Bradstreet's Hoovers subsidiary. Id. at 71. Based on the results of the study, NERA provided recommendations to Houston regarding the MWBE program. See Dkt. 32, Ex. 1.1A-1.
On May 1, 2014, Kossman filed this action, seeking to invalidate the MWBE program on the ground that it denies non-MWBEs equal protection of the law. Dkt. 1 at 1. Kossman is a company engaged in the business of providing erosion control services and is majority owned by a white male. Id. at 2. Kossman asserts that it has lost business as a result of the MWBE program because prime contractors are unwilling to subcontract work to a non-MWBE firm like Kossman. Id. at 6-7.
On July 24, 2015, Kossman filed a motion for summary judgment. Dkt. 22. On November 10, 2015, Houston filed a motion to exclude the testimony of Kossman's expert John Sullivan. Dkt. 30. On November 12, 2015, Houston filed a motion for summary judgment. Dkt. 32. On December 17, 2015, the court referred these motions to the Magistrate Judge. Dkt. 39.
On February 17, 2016, the Magistrate Judge issued the M&R. Dkt. 43. The Magistrate Judge found that Houston's motion to exclude Kossman's expert John Sullivan should be granted because Sullivan articulated no method and had no training in statistics or economics that would allow him to comment on the validity of the NERA study. Id. at 31-34. After a thorough analysis of the evidence presented, the Magistrate Judge found that the MWBE program was constitutional under strict scrutiny, except with respect to Native-American-owned businesses. Id. at 49-62. The Magistrate Judge found that there was insufficient evidence to establish a need for remedial action for businesses owned by Native Americans. Id. at 56.
On February 29, 2016, Kossman filed objections to the M&R. Dkt. 45. On March 10, 2016, Houston responded to Kossman's objections. Dkt. 46. Houston did not file its own objections to the M&R.
II. LEGAL STANDARD
For dispositive matters, the court "determine[s] de novo any part of the magistrate judge's disposition that has been properly objected to." Fed. R. Civ. P. 72(b)(3). "The district judge may accept, reject, or modify the recommended disposition; receive further evidence; or return the matter to the magistrate judge with instructions." Id. "When no timely objection is filed, the court need only satisfy itself that there is no clear error on the face of the record in order to accept the recommendation." Fed. R. Civ. P. 72(b), Advisory Comm. Note (1983). For nondispositive matters, the court may set aside the magistrate's order only to the extent that it is "clearly erroneous or contrary to law." Fed. R. Civ. P. 72(a).
III. OBJECTIONS
Kossman has filed the following objections to the M&R: (1) Houston improperly withheld the data underlying the NERA study; (2) the NERA study is based on insufficient, unverified information; (3) the NERA study uses unreliable anecdotal evidence; (4) NERA relied on stale data; and (5) the MWBE program cannot survive strict scrutiny because it is not narrowly tailored. Dkt. 45. The court will address Kossman's objections in turn.
A. Houston Improperly Withheld the Underlying Data
Kossman concedes that, based on the record to date, the Magistrate Judge correctly determined that Kossman's proposed expert John Sullivan articulates no method and relies on untested hypotheses. Id. at 4. Kossman acknowledges that Sullivan was unable to produce data to confront the NERA study. Id. However, Kossman argues that it was unable to conduct its own study because a study would need to be based on internal and confidential information that only Houston can access. Id. Kossman alleges that Houston wrongly withheld the data underlying the NERA study. Id. at 8-10. Houston claimed that NERA received the underlying data from Dun & Bradstreet, and NERA agreed it would not provide the Dun & Bradstreet data to Houston or any other party. Id. at 6. According to Kossman, it was later revealed that Houston's representation was false and that Houston's contract with NERA gave Houston access to the Dun & Bradstreet data. Id. at 7-10. Kossman apparently came to this conclusion based on an amicus brief filed by an Ohio attorney that was rejected by the Magistrate Judge. Id. at 6.
Houston's contract with NERA indicates that Houston "shall own without limitation all Documents that [NERA] prepares specifically and exclusively for the use of [Houston] under this Agreement." Id. at 7. However, the data was prepared by Dun & Bradstreet, not by NERA. NERA prepared the disparity study and recommendations for the MWBE program, which Houston produced along with many other documents NERA prepared in connection with the study. Dkt. 46 at 8-9. The court finds that this contract provision does not give Houston access to the Dun & Bradstreet data. Moreover, NERA's licensing agreement with Dun & Bradstreet prohibits NERA from providing the Dun & Bradstreet data to any third party, and this agreement apparently reflects NERA's standard policy. Dkt. 30, Ex. 2 at 233; Dkt. 45 at 6. It is unlikely that NERA would agree, on the one hand, to keep the Dun & Bradstreet data confidential, and on the other hand, to provide that data to Houston. Finally, the court agrees with Houston that Kossman would not be able to offer admissible analysis of the Dun & Bradstreet data even if it had access to the data. Dkt. 46 at 9. As the Magistrate Judge pointed out, Kossman's expert, Sullivan, is merely a legal scholar. Dkt. 43 at 30. Sullivan has no training in statistics or economics and would not be qualified to interpret the Dun & Bradstreet data or challenge NERA's methods. Id. at 31-32; see Wilson v. Woods, 163 F.3d 935, 937 (5th Cir. 1999) ("A district court should refuse to allow an expert witness to testify if it finds that the witness is not qualified to testify in a particular field or on a given subject."). Therefore, Kossman's objection is OVERRULED.
B. The NERA Study is Based on Unverified and Insufficient Information
Kossman argues that NERA relied on unverified information in conducting its study. Dkt. 45 at 5. Kossman notes that the NERA study makes explicit disclaimers of its reliability, including that the "[i]nformation furnished by others, upon which portions of this report are based, is believed to be reliable but has not been verified. No warranty is given as to the accuracy of such information." Id. Further, Kossman contends that the Dun & Bradstreet data provides an insufficient measurement of which firms are available to contract with the government. Id. at 12. Kossman maintains that bidding data is a superior measure of availability. Id. at 14.
It is unsurprising that NERA will not guarantee the information it relied on in creating its study and recommendations. NERA could not do so. NERA did not collect the data; Dun & Bradstreet collected the data. NERA's role was to analyze that data and make recommendations based on that analysis. NERA had no reason to doubt the authenticity or accuracy of the Dun & Bradstreet data, nor has Kossman presented any evidence that would call that data into question. As Houston has pointed out, Dun & Bradstreet data is extremely reliable, is frequently used in disparity studies, and has been consistently accepted by courts throughout the country. Dkt. 46 at 5-6 (collecting cases). Kossman has presented no evidence indicating that bidding data is a comparably more accurate indicator of availability than the Dun & Bradstreet data—rather, Kossman has relied on pure argument. Moreover, the court agrees with Houston and the Magistrate Judge that bidding data is inherently problematic because it reflects only those firms actually solicited for bids. Dkt. 43 at 54; Dkt. 46 at 11-12. Therefore, the bidding data would fail to identify those firms that were not solicited for bids due to discrimination, effectively undermining the purpose of the NERA study and the MWBE program. Kossman's objection is OVERRULED.
C. The Anecdotal Evidence is Unreliable
Kossman argues that NERA and Houston improperly relied on anecdotal evidence of discrimination. Dkt. 45 at 15-18. Kossman contends that anecdotal evidence is unreliable and unverified. Id.
Anecdotal evidence is a valid supplement to NERA's sophisticated statistical study. The case would be different if Houston had relied solely on anecdotal evidence. Here, however, Houston's MWBE program is supported by both statistical and anecdotal evidence. As the Magistrate Judge pointed out, anecdotal evidence provides a valuable narrative perspective that statistics alone cannot provide. Dkt. 43 at 58-59; see Coral Constr. Co. v. King Cty., 941 F.2d 910, 919 (9th Cir. 1991) (explaining that "the combination of convincing anecdotal and statistical evidence is potent," given that statistical evidence alone may not account for the complex motivations driving employment decisions). Houston was not required to independently verify the anecdotes. See Concrete Works of Colo., Inc. v. City & Cty. of Denver, 321 F.3d 950, 989 (10th Cir. 2003) (rejecting plaintiff's argument that anecdotal evidence of race or gender discrimination must be verified, and explaining that " [a]necdotal evidence is nothing more than a witness['s] narrative of an incident told from the witness['s] perspective and including the witness['s] perceptions"). Kossman could have presented contrary evidence, but it did not. Id. ("Denver was not required to present corroborating evidence and [plaintiff] was free to present its own witnesses to either refute the incidents described by Denver's witnesses or to relate their own perceptions on discrimination in the Denver construction industry."). Kossman's objection is OVERRULED.
D. NERA Relied on Stale Data
Kossman argues that NERA relied on data that is too old and therefore is no longer relevant. Dkt. 45 at 18-19. Kossman asserts that NERA relied on "regressions from 1993" and "data from nearly a quarter century ago." Id. at 19. The court finds that the data was not stale and was far more recent than Kossman would suggest. NERA used the most current available data at the time of the study. Dkt. 32, Ex. 1.1 at 32-34 (explaining that NERA relied on the most recent data available at the time the NERA study was undertaken, including Census Bureau data from 2006-2008, Federal Reserve data as recent as 2003, and surveys conducted between 1999 and 2007). The fact that NERA also considered older data alongside the more recent data does not undermine NERA's conclusions. Id. at 33 (explaining that NERA performed regression analyses on Federal Reserve data from 1993, 1998, and 2003, the three most recent waves of data). If anything, the ability to view and compare trends over time would seem to strengthen NERA's conclusions. Kossman has presented no evidence to suggest that NERA could have accessed more recent data or that NERA would have reached different conclusions with more recent data. Kossman's objection is OVERRULED.
E. The MWBE Program is not Narrowly Tailored
Kossman argues that the MWBE program is not narrowly tailored because Houston has failed to show that race neutral alternatives have been tried, evaluated, and found insufficient. Dkt. 45 at 19-20. Further, Kossman contends that the program places an undue burden on specialty trade contractors like Kossman who work in more specialized areas of large construction projects. Id. at 21-22.
The court agrees with the Magistrate Judge that the NERA study provides substantial evidence that race-neutral alternatives were insufficient. Dkt. 43 at 60-61. Despite the race-neutral alternatives in place in Houston, NERA consistently observed adverse disparities for MWBEs. Dkt. 32, Ex. 1.1A at 190-92; see also Dkt. 46 at 18 (cataloging the race and gender neutral measures in place in the Houston marketplace). Therefore, there is strong evidence that a remedial program was necessary to address discrimination against MWBEs. Further, Houston was not required to exhaust every possible race-neutral alternative before instituting the MWBE program. See Grutter v. Bollinger, 539 U.S. 306, 339, 123 S. Ct. 2325 (2003) ("Narrow tailoring does not require exhaustion of every conceivable race-neutral alternative.").
The court also finds that the MWBE program does not place an undue burden on Kossman or similarly situated companies. Under the MWBE program, a prime contractor may substitute a small business enterprise like Kossman for an MWBE on a race and gender neutral basis for up to 4% of the value of a contract. Dkt. 32, Ex. 3 at 3. As Houston has pointed out, Kossman has not presented evidence that it ever bid on more than 4% of a Houston contract. Dkt. 46 at 20. Moreover, the fact that the MWBE program may place some burden on Kossman is insufficient to support the conclusion that the program is not narrowly tailored. As the Magistrate Judge correctly observed, "the proportional sharing of opportunities is, at the core, the point of a remedial program." Dkt. 43 at 62. The court agrees with the Magistrate Judge's conclusion that the MWBE program is narrowly tailored. Therefore, Kossman's objection is OVERRULED.
IV. NATIVE-AMERICAN-OWNED BUSINESSES
The NERA study found that Native-American-owned businesses were utilized at a higher rate in Houston's construction contracts than would be anticipated based on their rate of availability in the relevant market area. Dkt. 32, Ex. 1.1A at 191. This finding would tend to negate the presence of discrimination against Native Americans in Houston's construction industry. However, NERA noted that the results for Native Americans should be interpreted with caution. Id. at 190. NERA explained that the high utilization rate for Native Americans stems largely from the work of two Native-American-owned firms. Id. Without these two firms, the utilization rate for Native Americans would decline significantly, yielding a statistically significant disparity ratio. Id.
The Magistrate Judge found that there was insufficient evidence to support including Native Americans in the MWBE program. Dkt. 43 at 56. The Magistrate Judge explained:
NERA suggested that a significant statistical disparity would exist if two of the contracting Native-American-owned businesses were disregarded. That opinion is not evidence of the need for remedial action. The court finds no equal-protection significance to the fact that the majority of contracts let to Native-American-owned businesses were to only two firms. The utilization goal for businesses owned by Native Americans is not supported by a strong evidentiary basis.Id. Therefore, the Magistrate Judge recommended that the court grant summary judgment in favor of Kossman with respect to the utilization goal for Native-American-owned businesses. Id. at 63.
Houston did not object to this recommendation. When a party does not file a timely objection to a magistrate judge's recommendation on a dispositive motion, the court reviews the recommendation for clear error only. Fed. R. Civ. P. 72(b), Advisory Comm. Note (1983). In the M&R, the Magistrate Judge correctly considered whether there was an evidentiary basis to support the inclusion of each minority group in the MWBE program. See Contractors Ass'n of E. Pa., Inc. v. City of Phila., 6 F.3d 990, 1007-08 (3d Cir. 1993) (explaining that strict scrutiny requires the court to determine whether there is evidence of discrimination as to each minority group included in a comparable city ordinance, and finding insufficient evidence to support inclusion of businesses owned by Hispanic, Asian-American, or Native-American persons). The NERA study provided some evidence that Native Americans faced discrimination in the construction industry. See, e.g., Dkt. 32, Ex. 1.1A at 102 (finding a statistically significant disparity in business owner earnings in the construction sector among Native-American-owned businesses). However, given that Native-American-owned businesses were utilized at a higher rate in Houston's construction contracts than their availability rate would predict, the court cannot say that the Magistrate Judge's recommendation is clearly erroneous. Further, as the Magistrate Judge concluded, there is limited significance to the fact that utilization of Native-American-owned businesses would drop to statistically significant levels if two Native-American-owned businesses were ignored. The situation presented by the hypothetical non-existence of these firms is not evidence and cannot satisfy strict scrutiny. Therefore, the court ADOPTS the Magistrate Judge's recommendation with respect to the utilization goal for Native-American-owned businesses. However, the court notes that a preference for Native-American-owned businesses could become constitutionally valid in the future if there were sufficient evidence of discrimination against Native-American-owned businesses in Houston's construction contracts. See Contractors Ass'n of E. Pa., 6. F.3d at 1008 ("Nothing in our decision prevents the City from re-enacting a preference for construction firms owned by Hispanic, Asian-American, or Native American persons based on more concrete evidence of discrimination.").
V. CONCLUSION
Kossman's objections (Dkt. 45) are OVERRULED, and the M&R (Dkt. 43) is ADOPTED in full. Houston's motion to exclude (Dkt. 30) is GRANTED. Kossman's motion for summary judgment (Dkt. 22) is GRANTED with respect to the utilization goal for Native-American-owned businesses and DENIED in all other respects. Houston's motion for summary judgment (Dkt. 32) is DENIED with respect to the utilization goal for Native-American-owned businesses and GRANTED in all other respects. A final judgment will issue consistent with this order.
Signed at Houston, Texas on March 22, 2016.
/s/_________
Gray H. Miller
United States District Judge