White Paper: Doubts About INDOT Disparity Study
The author is an attorney with a national practice specializing in disparity studies, called as an expert witness in challenges to racial preferences in public contracting. He wrote this for the foundation based on work being commissioned by a client group of contractors. Contact him at email@example.com. The foundation’s white papers are intended to make scholarly research on Indiana issues more widely available to policy analysts and researchers. White papers represent research in progress and are published to invite comment and discussion as preparation for their submission to academic journals and other professional publications. The authors are solely responsible for the content of their research and analysis.
by JOHN SULLIVAN, J.D.
(Sept. 10) — The Indiana Department of Transportation (INDOT) has released its 2015-2016 disparity study intended to serve as the justification for and the statistical source of INDOT’s Disadvantaged Business Enterprise (DBE) goals and its Minority and Women owned Business Enterprise (MEBE) goals.
Conducted by BBC Research & Consulting of Denver, the study covers contracts awarded by the agency between July 1, 2009, and June 30, 2013. The firm reviewed contracts in construction; professional services, goods and support services. This article focuses on construction, which represents 87 percent of all contracting dollars awarded by INDOT during the period covered by the study.1
Does the INDOT disparity study supply the needed support for the agency’s contracting preferences? Contracting preferences based on race, gender and ethnicity, like those INDOT imposes on its contracts, are constitutional only if the evidence supporting those preferences meets the standards of City of Richmond vs. Croson.2 In Croson, the Supreme Court decided that a local government or agency could not implement race conscious programs to benefit minority groups unless the government had a “strong basis in evidence for its conclusion that remedial action was necessary.”3
If no such strong basis in evidence of discrimination has been shown in the INDOT marketplace, then the agency’s DBE and MWBE programs serve no remedial purpose and so are not constitutionally permissible. The INDOT disparity study is the agency’s attempt to establish a strong basis in evidence of discrimination.
A disparity study is, at its core, an analysis of whether minority- and women-owned firms have received a smaller percentage of government contracting dollars than expected in light of their availability in the marketplace. These statistical disparities are expressed in terms of disparity ratios comparing a group’s availability with its utilization. Disparity ratios must also control for nondiscriminatory variables which could affect the outcome.
A properly conducted disparity study availability analysis identifies those firms in the marketplace that are “qualified, willing and able” to perform the government’s contracts. As Croson said:
“Where 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 or the locality’s contractors, an inference of discriminatory exclusion could arise.”4
The INDOT disparity study properly recognizes the importance of availability computations: “The purpose of the availability analysis was to provide precise and representative estimates of the percentage of INDOT contracting dollars for which minority- and woman-owned businesses are available.”5 The BBC study for INDOT is massive, nearly 600 pages. In all these pages has the BBC study succeeded in providing availability MWBE percentages that are “precise”? For all INDOT contracts during the 2009-2013 time frame, the BBC study found overall MWBE availability of 11.5 percent. Of that, 8.2 percent of the firms were owned by white women.6 When INDOT construction is looked at separately, overall ability is 11.0 percent, of which 8.6 percent were white woman-owned.7
The way BBC reached those percentages is complex. The availability methodology starts with potential availability. The initial steps toward computing potentially available firms are not controversial: determine the proper market area (Indiana) and the sub-industries in which INDOT contracts.8
The next step, however, is problematic: “The study team then developed a database of potentially available businesses through surveys with businesses” in Indiana in the relevant sub industries. The database of potentially available firms was developed through a survey asking questions intended to determine qualifications for and willingness to do public contracting, as well as capacity of a firm to perform large contracts.9
The survey results were then used as part of the study’s “contract-by-contract matching approach to calculate availability.” This matching process looked at factors of the prime or subcontract awarded by INDOT (type of work, location, contract size, etc.) and “BBC then identified businesses in the availability database” of potentially available firms that could perform that contract.10
Certainly the telephone survey is central to the BBC study’s approach to availability. There are many problems with this. For starters, most persons the survey attempted to contact could not even be contacted. And of those contacted, only 2,200 (out of 27,384 the study attempted to contact) actually completed the interview about their firm’s characteristics.11
Even overlooking the low response rate, there are crippling problems with the use of a survey in this way. Survey recipients were asked: “Is your company qualified and interested in working with state or local agencies in Indiana as a subcontractor, trucker/hauler or supplier?”12 The problem with basing potential availability on answers given to this question in a telephone survey is that whoever answered the survey was free to interpret the term “qualified” however that person desired.
Qualifications are of course a matter of great significance in public contracting. Otherwise, awarding contracts solely on the basis of low bid might result in shoddy work or cost overruns. INDOT and the people of Indiana do not want the work on a bridge, highway or tunnel done by an unqualified construction company. But a firm calling itself qualified does not make that firm qualified. The INDOT study assumes that it does.
There’s a similar problem with the survey asking about a firm’s willingness to contract with INDOT.13 What could it hurt a firm to reply, “Sure we’re interested in working” on an INDOT project? No effort to actually compete for an INDOT construction contract was a prerequisite for being deemed willing to compete for INDOT contracts. Claiming willingness was enough.
Most of the answers to the surveys came from telephone interviews in which information was supplied by individuals as varied as CEOs, receptionists, sales managers and presidents. Would a receptionist necessarily know accurate answers to questions like, “What was the largest contract or subcontract that your company bid on or submitted quotes for during the past five years?”14
Certainly there could have been verification on this question of bidding or supplying quotes. Whether a firm actually bid on a contract is a matter of public record.
A firm taking the time and money to bid on a contract can be viewed as qualified, willing and able to perform the contract. A construction contractor has the opportunity to bid on hundreds of jobs each year, many more than he has the time to estimate or the capacity to perform. And, since bidding a job necessarily entails considerable expense in both time and money, it is imperative that jobs be selected that offer at least a fair chance of earning a profit. Consequently, bidding firms are truly available firms. But there is no indication that the BBC study made any attempt to examine contract bidders.15
Another legitimate concern unaddressed by the study is that whoever answered the phone might not necessarily want to reply with accurate information to a question like, “Roughly, what was the average gross revenue of your company, just considering your location, over the last three years?”16 Answers were accepted as accurate. There was no external verification of the survey answers given.
The BBC study concluded “that the availability of potential DBEs for INDOT’s FHWA-funded transportation contracts is 10.8 percent.”17 This figure is almost exactly the percentage INDOT is proposing as its DBE goal for federal fiscal years 2017-2019. The proposed goal is 10.9 percent.
The BBC study looked at a total of 17,079 construction contracts awarded by INDOT during the four-year time frame. These construction contracts had an aggregate value of $5,543,480,000. Of these contracts, 13,927 were federally funded at least in part; the remaining 3,152 contracts were entirely state-funded. The federally funded contracts were much larger, accounting for $4.7 billion compared with $782 million in wholly state-funded construction contracts. These figures represent prime and subcontracts combined.18
Many disparity studies have trouble obtaining complete subcontracting data and the INDOT study is no exception. While prime contractor data was complete, the study admits that it was necessary “to gather more comprehensive subcontract data.” To do so, “the study team conducted surveys with prime contractors to collect information on the subcontracts” with INDOT.19 The study acknowledges that one-third of the dollars on these contracts was not captured, so a good amount of INDOT subcontracting dollars are not covered in the study. This is a critical problem since both MWBEs and DBEs do a large amount of their work on subcontracts. As a result of this problem, it may well be true that MWBE and DBE utilization has been understated, creating a disparity where perhaps one does not exist or exists to a smaller degree.
The study includes the information that a small number of DBEs/MWBEs received a disproportionate share of INDOT’s construction dollars. For instance, a single Native American water, sewer and utility-line contractor received $74 million of the $87 million awarded to Native American firms in the time frame.20
If the intent of the MWBE and DBE preferences is to expand the number of minority and woman-owned firms participating in INDOT contracting, the agency is not succeeding.
Every disparity study of course has disparity ratios. A disparity ratio shows the relationship between the utilization of DBE/M/WBE firms and their availability in the marketplace.
As the BBC study explains, a disparity ratio (which the study terms a “disparity index”) “provides a way of assessing how closely the actual participation” of MWBEs “matches the percentage of contract dollars that those businesses might be expected to receive based on their availability . . .”21
The disparity ratio results do not justify inclusion of all groups in INDOT’s preferences. This is true even when the study’s reliance on an unreliable survey is ignored.
The BBC study correctly concludes that, “Utilization and disparity analysis results, along with other pertinent information, are relevant to INDOT’s determination of which groups could be eligible for any race- or gender-conscious measures.”22 But the disparity study results do not justify the inclusion of all DBE groups, even when the flaws of the study’s survey are overlooked. To understand why, it is necessary to be aware of Croson and lower-court decisions that have established the requirement there be sufficient justification for every group benefitting from preferences in public contracting.
In Croson one of the rationales on which the Supreme Court relied to invalidate Richmond’s public-contracting preferences was the program’s overly expansive definition of minority group members, which encompassed citizens “of the United States who are Blacks, Spanish-speaking, Orientals, Indians, Eskimos and Aleuts.”23 The Court criticized the inclusion of racial groups which, as a practical matter, may never have suffered from discrimination in the local construction marketplace. Evidence of local discrimination against each group was needed for that group’s inclusion in the MBE program at issue in Croson.
Numerous circuit courts have held that in order for a preferential program to be narrowly tailored there must be a showing of discrimination as to each group included in the preferential program. Western States Paving Company vs. Washington State Department of Transportation,24 is typical. In that case the Ninth Circuit required that each of the groups benefitting from a preferential program must have suffered discrimination within the state. Otherwise, the program “provides minorities who have not encountered discriminatory barriers with an unconstitutional advantage.” The Circuit explained that, “a remedial program is only narrowly tailored if its application is limited to those minority groups that have actually suffered discrimination.”
Accordingly, each of the minority groups benefitted by INDOT’s MWBE and DBE program must be shown to have experienced discrimination within the Indiana construction marketplace. The disparity ratios in the BBC study not only fail to establish the presence of discrimination against all the DBE/MWBE groups benefitting from the agency’s program, the study does not even show statistically significant underutilization for those groups. Without statistically significant underutilization for a DBE/MWBE group, that group should not be the beneficiary of a preference.
The disparity ratio results for all INDOT contracts show that two groups are actually overutilized – that is, the group receives more than its expected share of contracting dollars. These groups are Asian Pacific-owned firms, which received three and a half times their expected share and Native American-owned firms, which got six and a half times their expected share. According to the group-specific requirement of Croson and lower courts applying Croson, preferences for these groups would not be justified.
On INDOT construction contracts during the time frame, MWBEs as a whole were overutilized. When groups are considered separately on construction contracts, Blacks, Asian Pacific Americans and Native Americans were overutilized, and not by a little. Asian Pacific-owned construction firms received six times their expected utilization and Native American-owned construction companies got eight times their expected utilization. Subcontinent Asian Americans and Hispanic Americans were underutilized. White women were underutilized, but not enough to meet the Croson standard of statistically significant underutilization.25 Under the group-specific requirement, construction firms owned by Blacks, Asian Pacific Americans, Native Americans and white women would not be eligible for preferences. Only Subcontinent Asian- and Hispanic-owned construction companies would be.
What theory of discrimination in public construction contracting could possibly explain these patchwork results?
Where Is the Claimed Discrimination Occurring?
Prime construction bids are opened in public. INDOT construction contracts are awarded to the lowest responsible bidder, as the BBC study makes clear: “The Construction Division (of INDOT) awards all contracts on a low-bid basis, and there are no dollar thresholds above or below which procurement policies change.”26 There is no subjectivity in the award of agency construction contracts. Without subjectivity, how can discrimination take place in the award of INDOT prime contracts?
Nor has this disparity study shown discrimination against MWBE and DBE subcontractors. To the contrary, the study reveals that as subs MWBEs as a whole are not underutilized but greatly over utilized. When groups are considered separately, white women, Blacks, Asian Pacific Americans and Native Americans are overutilized on construction subcontracts. Certainly prime contractors are not discriminating against MWBEs and DBEs if these groups receive far more than their anticipated subcontracting share. Only Subcontinent Asians and Hispanics are underutilized as subs.27 No theory of discrimination could explain these results.
When all federally funded construction prime and subcontracts are considered together, it is white male-owned construction businesses that are underutilized.28 With no discrimination alleged on the prime contracting level and DBE subcontractors overutilized and white males underutilized, the question remains: Where is the discrimination, if any, taking place? The study offers no persuasive answer.
Contracts Without DBE Goals
The BBC study admits that, “Disparity analysis results indicated that most racial/ethnic and gender groups did not show disparities on contracts to which INDOT applied DBE of MWBE contact goals during the study period.” The study, however, went on to say, “In contrast, most racial/ethnic and gender groups showed substantial disparities on contracts to which INDOT did not apply DBE or MWBE goals.”29 That statement is not nearly true for Hispanics, which received almost one-and-a-half times their expected share of contracting dollars.30
With no discrimination shown on the disparity ratios for most DBE groups in construction, and without the disparity ratios showing where in the INDOT public construction contracting process discrimination is happening, it is here on contracts without DBE goals that INDOT justifies its proposed DBE goals for all six groups: Black American-owned businesses; Asian Pacific American-owned businesses; Native American-owned businesses; Subcontinent Asian American-owned businesses; and white woman-owned businesses.31
Figure 5 of the agency’s proposed DBE goal setting report provides disparity ratios for INDOT contracts for which no goals were applied.32 On these disparity ratios each MWBE/DBE group shows significant disparity, except for Hispanics. The goal-setting report claimed Hispanics deserved to be included in INDOT’s program since, “Hispanic American-owned businesses showed substantial disparities for several other key contract sets including all INDOT contracts considered together (disparity index of 54) and on INDOT’s FHWA-funded contracts (disparity index of 73).”
It is unsurprising that typically DBEs receive a larger percentage of subcontracting dollars on contracts with goals than on those contracts without goals. If a prime does not meet the MWBE or DBE goal, that prime risks losing the project — unless good faith efforts can be shown. The motivation to meet the goal and avoid the risk of a low-bidding prime losing the work is obvious. This pattern does not establish discrimination. What it does is illustrate the power of imposing goals on public contracts.
The study then recommends the expansion of goals set on contracts without asking why the disparity between goaled and non-goaled contracts currently exists.
Annual DBE Goal-Setting For DBES and MWBES
Having found justification for DBE goals for all groups, the study turns to determining the actual annual goals for MWBEs and DBEs. The BBC study explains that each year a state MBW/WBE Commission “establishes overall annual goals” for MWBEs. A goal for construction is set separately from goals set for professional services or goods and support services. Goals apply on both the prime and subcontracting level. The study insists the MWBE goals are aspirational, though construction contractors might say differently. The study also points out that, “There are no provisions in the MBE/WBE program that give explicit preferences to MWBEs over majority-owned businesses.”
However, the study also details how prime contractors must either meet the MWBE goal or be granted a good-faith waiver.33 The impact of the MWBE goal on whether a low-bidding prime gets the contact is apparent.
The study reaches an availability percentage for MWBEs of 11.5 percent.34 This figure is not specific to construction but is for INDOT contracts in professional services, goods and support services as well as construction. Here are the group-specific Step One availability percentages for federally funded construction (based on total DBE availability of 10.8 percent):
Anecdotes: Qualitative Information About Marketplace Conditions
Anecdotal allegations of discrimination are part of almost every disparity study, as they are part of the BBC INDOT study. No race-conscious program has ever been upheld solely on the basis of anecdotes. A typical court decision on anecdotes is Coral Construction vs. King County. In that case, the Ninth Circuit explained the limited effectiveness of anecdotes:
“While anecdotal evidence may suffice to prove individual claims of discrimination, rarely, if ever, can such evidence show a pattern of discrimination necessary for the adoption of an affirmative action plan … the MBE program cannot standard without a proper statistical foundation.”35
The anecdotes in the BBC disparity study would certainly not be sufficient to justify the INDOT MWBE or DBE program. The anecdotes — what the BBC termed “Qualitative information about marketplace conditions” — were obtained through input from “more than 600 business and trade representatives.”
This input came in four forms: in-depth interviews; availability surveys; public forums; and written testimony. “The study team conducted in-depth interviews with 71 businesses and four trade associations.” These interviews touched on a number of topics, including “experiences working or attempting to work with Indiana state entities.” Interview participants were selected by a random sample of businesses stratified by business type, location and race/ethnicity/gender of the owner. Some of those interviewed worked as primes, some worked as subs and some worked as both.
A total of 555 businesses provided comments to the availability surveys, which asked firm owners and managers whether their companies have experienced barriers starting or expanding businesses. There were three public forums held around the state in January 2015 in Gary, Indianapolis and Evansville. Written testimony was received through e-mail.36
A number of conclusions can be drawn from the many anecdotal responses paraphrased or quoted in the lengthy Appendix E. Reading the many pages of Appendix E, it becomes apparent that many problems are by no means limited to MWBEs and DBEs. For example, a “non-Hispanic white male owner of a construction firm reported the lack of a consumer base is the biggest problem faced for new companies.” 37
Many of those whose responses appear in the chapter view slow payment as a problem. As “The non-Hispanic white female owner of a WBE-certified goods and services firm indicated . . . slow payment causes a problem for all small businesses.” Learning about subcontracting opportunities was also a problem for many.38 The solution to that is better communication to all potential subcontractors, regardless of the race, ethnicity, or gender of the company owner.
It is instructive to see comments in Appendix E comparing the experiences of businesses in the public sector with the private sector. For example: “Public work is all low-bid. So, we are forced to (use) the lowest subcontractors, which sometimes affects scheduling and quality of work. Private sector work is based on quality as well as price. We can chose our subcontractors” and consequently produce better work.”39
There are a few claims of discrimination in the anecdotal section of the disparity study. ”A Native American male owner of a non-certified construction firm stated that while he cannot prove any allegations of discriminatory treatment, he feels it exists.” And, “A Hispanic American female owner of an MBE- and WBE-certified professional services firm expressed that she believes there are additional barriers to being a woman- and minority-owned business.”40
It is vital to note that no attempt at verifying these claims of discrimination has been made. There are two reasons why anecdotes should be investigated for accuracy:
- It seems a matter of basic fairness that any preferential program which disadvantages some people according to their race, ethnicity and gender while advantaging others — as the INDOT DBE and MWBE programs do — should only be permitted where discrimination has been shown to be real, and not simply perceived. Unless they are investigated, these anecdotal accounts remain mere perceptions of discrimination.
- Unless the claims of discrimination can be verified, the right remedy to that discrimination cannot be fashioned. To find and implement the most effective remedy, perception must be distinguished from reality. For example: What if an MBE interviewee claimed he was denied payment by a prime due to discrimination — when investigation of the claim would have revealed that the real explanation was incomplete and sloppy work as a sub? This individual’s claim could have been verified or disproven. If discrimination did indeed occur, those responsible could be debarred from INDOT contracting. That would solve the problem; but a sanction such as debarment would not be justified unless the problem actually happened, and was not merely perceived.
The 2015-2016 disparity study done by BBC for the Indiana Department of Transportation is the basis for MWBE and DBE goals the agency imposes on construction contracts. Is the study sufficiently persuasive to justify the proposed DBE goals? It seems unlikely:
- At the heart of any disparity study is its approach to availability, and at the heart of the BBC study’s approach to availability is a telephone survey. The survey may well have been answered by people unwilling to give accurate answers. Or the survey respondent may have said that sure, their firm was qualified and willing to contract with INDOT, when in reality that was not true. We will never know, since there was little verification of the answers given.
- Subcontractor utilization data was incomplete. Since it is on the subcontracting level that MWBEs and DBEs do so much of their work this could be a crippling flaw.
- Croson and subsequent court decisions require that there be a showing of significant underutilization for each DBE group and that has not been shown in the BBC study.
- The disparity ratios in the study do not indicate where discrimination is occurring. The study fails to show discrimination on either the prime contracting or subcontracting level.
- The study’s lengthy anecdotal section reveals that may of the problems cited are not examples of discriminatory behavior but problems faced by all small businesses. None of the anecdotal claims were investigated for accuracy, as they should have been.
What part of the study, then, is support for the proposed DBE goals? The goals are supposedly supported by “the substantial disparities on contracts to which INDOT did not apply DBE or MWBE goals.”41 The problem with that reasoning is that it assumes these disparities are caused by discrimination.
In fact, the coercive nature of DBE goals — either meet the goals or show good faith effort or lose the contract — is the likely explanation for these disparities.
1. Proposed INDOT DBE goal setting report, p. 4.
2. Croson, 488 US 469 (1989).
3. Croson, 488 US at 500.
4. Croson, 488 US at 509.
5. INDOT study, Appendix C, p. 14.
6. INDOT study, Figure F-2.
7. INDOT study, Figure F-3.
8. INDOT study, Chapter 5, p. 1.
9. INDOT study, Chapter 5, pp. 1-3.
10. INDOT, chapter 5, p. 4.
11. INDOT study, Appendix C, pp. 11-12.
12. INDOT study, question B10, Appendix C, p. 23
13. INDOT study, question B10. Appendix C, p. 23.
14. INDOT study, Appendix C, pp. 29, 33.
15. INDOT study, Appendix C, p. 17.
16. INDOT study, Appendix C, p. 32.
17. “Information Regarding the Disadvantaged Business Enterprise Goal for Federal Fiscal Years 2017 through 2019,” p. 4.
18. INDOT study, Chapter 4, Figure 6-1; p. 6.
19. INDOT study, Chapter 4, p. 5.
20. INDOT study, Chapter 6, pp. 3-4.
21. INDOT study, Chapter 7, Figure 7-1; p. 1.
22. INDOT study, Executive Summary, p. 3.
23. Croson, 488 US at 478.
24. 407 F.3d 983, 999 (2005).
25. INDOT study, Chapter 8, Figure 8-2; p. 3.
26. INDOT study, Chapter 4, p. 1.
27. INDOT study, Chapter 8, Figure 8-3; pp. 2-3.
28. INDOT study, Appendix F, Figure F-20.
29. INDOT study, Executive Summary, p. 8.
30. INDOT study, Chapter 7, p. 7; Figure F-16;
31. INDOT proposed DBE goal setting report, p.10.
32. INDOT proposed DBE goal setting report, p. 9.
33. INDOT study, Chapter 1, pp. 2-3.
34. INDOT study, Chapter 5, p. 7.
35. 941 F.2d 910, 919 (1991).
36. INDOT study, Appendix E, pp. 1-3.
37. INDOT study, Appendix E, p. 6.
38. INDOT study, Appendix E, pp. 53; 78.
39. INDOT study, Appendix E, p. 28.
40. INDOT study, Appendix E, pp. 10; 128.
41. INDOT study, p. 8.