Finding Nothing
An important report involving the Adarand case.

By George R. La Noue, professor of political science, University of Maryland Baltimore County
October 31, 2001 8:55 a.m.

oday, in Adarand v. Mineta, the Supreme Court will again confront the contentious issue of racial preferences in public contracting. The outcome of this case may not only determine the future of federal affirmative-action programs, but influence state and local programs as well. Not surprisingly, the case has attracted more than a dozen amicus briefs, and both sides have lodged thousands of pages of documents with the Court. The most influential document may be a recent report issued by the General Accounting Office (GAO) which substantially undermines the government's effort to defend racial preferences.

GAO is frequently called the congressional watchdog; its most frequent assignment is investigating bureaucratic behavior at the behest of individual congressmen or committees. Thanks to the GAO's 3,500 employees and the scores of reports they write annually, the federal government in its mundane daily operations is more responsive, equitable, and efficient than it otherwise might be.

This June, however, GAO issued a report of enormous constitutional significance because of what it didn't find, and because it challenged a number of Congressional assumptions when the transportation racial-preference program was extended.

After a three-year investigation, GAO's report — "Disadvantaged Business Enterprises: Critical Information is Needed to Understand Program Impact" — did not find evidence of any pattern of discrimination against minority- and women-owned businesses (DBEs) in the highway construction industry. The report, issued just a few days before the plaintiff's brief was due in Adarand, was reproduced in full in an appendix to that document now before the Court. Since it is the government's burden to show that racial preferences are a response to identified discrimination, the GAO report now constitutes a major hurdle.

Further, the significance of the report is enhanced by its unusual origin. During the 1998 passage of a massive federal highway program, TEA-21, Republicans made major efforts to amend the two-decade-old DBE program to make it race-neutral, benefiting all small businesses. In the face of nearly unanimous opposition from Democrats, a compromise was reached to preserve the DBE program, while requiring GAO to study the characteristics of DBE firms and whether there was any evidence of discrimination against them. This was putting the evidentiary cart before the legislative horse, because courts have clearly said the proof of discrimination should exist before a narrowly tailored, race-conscious remedy is used. It was, however, the best Sen. McConnell (R., Ky.) and his colleagues could do. As Congressman Schuster (R., Pa.), chair of the House Committee on Transportation and Infrastructure and the floor manager for the transportation bill, said during the House debate, the Act "...also requires a GAO study that would examine whether there is continued evidence of discrimination against small businesses owned and controlled by socially and economically disadvantaged individuals. I believe such a study will lay the groundwork for future reform."

The GAO study has four major components: an attempt to describe the characteristics of DBEs and their non-DBE competitors, an examination of discrimination complaints filed by DBEs, a review of transportation-related "disparity studies," and an answer to what happens when DBE programs are discontinued.

1. Missing information

If DBEs are truly "disadvantaged," then it is essential to have information to compare them with non-DBEs. That information rarely exists. GAO had planned to survey all transit authorities receiving federal funds, but the Federal Transit Administration does not even have a complete list. When 83 state and transit recipients were surveyed, only 40 percent or fewer of the respondents could report the gross revenues of the DBEs that won contracts. Fewer than 25 percent of the respondents could report the gross revenues of the DBEs that did not win contracts. Only about a third of the agencies could report data on the personal net worth of DBE owners, although TEA-21 regulations require that such owners' net worth not exceed $750,000. Even fewer recipients could report data on the gross revenues or owner net worth characteristics of non-DBE firms. While most respondents could report data about subcontracts-awarded DBEs, fewer than a third could report similar data for non-DBEs. That means that most respondents did not regard comparing DBE and non-DBE subcontractor utilization as relevant in setting goals, or in determining whether discrimination exists.

Nor are respondents acquiring relevant information: 98.8 percent have not conducted any study determining if awarding prime or sub contracts to DBEs affects contract costs; 67.5 percent report no study of discrimination against DBE firms; 84.2 percent report no study of discrimination against DBEs by financial credit, insurance, or bond markets; 79.5 percent report no study of factors making it difficult for DBEs to compete; and 92.8 percent have made no study on the impact of the DBE program on competition and the creation of jobs.

2. Discrimination complaints

GAO conducted a survey of discrimination complaints received by USDOT and recipients, as that might prove to be a useful source of data about the nature and frequency of discrimination against DBEs. Although USDOT sometimes receives written complaints of discrimination, the agency could not supply information on the number of complaints filed or investigations launched, or their outcomes. GAO also asked state and local transit recipients about complaints they received, and they had better data. During 1999 and 2000, 81 percent of the recipients had no complaints, while a total of 31 complaints were received by the other recipients. Of these, 29 were investigated, and findings of discrimination were made only four times across the nation. The report concluded:

Other factors may also limit the ability of DBEs to compete for USDOT state-assisted contracts. The majority of states and transit districts we surveyed had not conducted any kind of analysis to identify these factors. Using anecdotal information, we identified a number of factors, or barriers, such as a lack of working capital and limited access to bonding, that may limit DBEs' ability to compete for contracts. However, there was little agreement among the officials we contacted on whether these factors were attributable to discrimination.

In fact, GAO reported there were few, if any, studies by government agencies or industry groups regarding barriers to DBE contracting. "USDOT officials, however, stated that they believe contract bundling is one of the largest barriers for DBEs in competing for transportation contracts." That, of course, is not a problem caused by discrimination.

3. Disparity studies

GAO also reviewed 14 transportation-specific disparity studies completed between 1996 and 2000, because they might provide evidence about discrimination and because USDOT permits recipients to use disparity studies to set annual goals, and determine the level of discrimination these goals purportedly are remedying. GAO found that about 30 percent of the recipients surveyed used disparity studies to set their 2000 goals.

GAO, however, found these studies were not reliable because the limited data used to calculate disparities, compounded by the methodological weaknesses, create uncertainties about the studies findings... While not all studies suffered from every problem, each suffered enough problems to make its findings questionable. We recognize there are difficulties inherent in conducting disparity studies and that such limitations are common to social science research; however, the studies we reviewed did not sufficiently address such problems or disclose their limitations.

USDOT in fact made this problem worse, because it advised recipients that disparity studies should be "reliable" and then provided no guidance on what would comprise a reliable study. GAO concluded that:

USDOT's guidance does not, for example, caution against using studies that contain the types of data and methodological problems that we identified above. Without explicit guidance on what makes a disparity study reliable, states and transit authorities risk using studies that may not provide accurate information in setting DBE goals.

GAO's finding about the unreliability of disparity studies is consistent with the findings of every court that has examined the merits of such studies after discovery and trial.

4. Discontinuing Programs

One of the arguments used in the TEA-21 debates and defendants' trial briefs is the assertion, often anecdotal, that without goals, DBE participation would decline precipitously. The difficulty with that assertion, even if true, is that the decline in DBE participation may be the result of previous overutilization caused by goals set too high, or may occur because when a program is struck down, DBEs may have little incentive to seek or maintain certification.

But is the basic assertion true? It turned out that ten of twelve recipients with discontinued programs did not know what the DBE participation result was. For instance, though Michigan was cited frequently by DBE proponents in the TEA-21 debate as an example where minority utilization declined after a program was struck down, GAO reports:

Michigan could not provide us with minority- and women-owned business participation data in state highway contracting for the years immediately before and after it discontinued its program. Furthermore, Michigan officials stated that the analysis showing the decline that is often cited was a one-time-only analysis and that analysis is no longer available. Consequently we cannot verify the number cited during the debate.

While GAO, of course, does not conclude that Congress did not have a compelling interest for the DBE program, the missing and/or unreliable data GAO discovered surely undermines the rhetorical assertions of DBE advocates. Nothing in the GAO report supports any finding of a national pattern of discrimination against DBEs. No doubt every law clerk on the Supreme Court will become familiar with this GAO report as the Adarand decision is crafted.