Section 26.45 How Do Recipients Set Overall Goals?
Since its inception, the recipient's overall goal has been the heart of the DBE program. Responding to Adarand, DOT clarified the theory and purpose of the overall goal in the SNPRM. In the proposed rule, the Department made clear that the purpose of the overall goal-- and, in fact, the DBE program as a whole--is to achieve a ``level playing field'' for DBEs seeking to participate in federal-aid transportation contracting. To reach a level playing field, recipients need to examine their programs and their markets and determine the amount of participation they would expect DBEs to achieve in the absence of discrimination and the effects of past discrimination. The focus of the goal section of the SNPRM was to propose ways to measure what a level playing field would look like and to seek input on the availability of data to make such a measurement.
The Proposed Rule and Comments
The Department proposed several options that recipients might use for setting overall goals, including three alternative formulas for measuring the availability of ready, willing and able DBEs in local markets. The specific formulas will be discussed below, but generally, they each called for setting a goal that reflected the percentage of locally available firms that were DBEs (i.e. dividing the number of DBEs by the number of all businesses). On all of the alternatives, the SNPRM sought comments on both the feasibility and practical value of the options, as well as the prospects for combining any of the approaches and the question of whether to mandate a single approach or allow each recipient to choose amongst the options. We invited commenters to propose changes to any of the details of the options or to devise entirely new ones. Finally, we asked commenters for their input on the availability of reliable data for use with each of the options.
Hundreds of commenters of all types--including DBEs and non-DBEs, prime and subcontractors, state and local recipients, industry and interest groups and private individuals--responded with a wealth of feedback, opinions and data. It is an understatement to say that there was no consensus among commenters as to the best way to set overall goals. Support for the proposed options was almost evenly spread over the choices presented, with many commenters firmly against all of the options. Still more suggested that the current, non-formulaic method was the best way to ensure the flexibility to respond to local market conditions. Similarly, among those who expressed an opinion, commenters were split between the propriety of choosing a single ``best'' method and imposing it on all recipients and allowing recipients to choose amongst all the options. One of the few universal themes in the goal- setting comments was the problem of the availability of reliable data on the number of DBE and non-DBE contractors.
There were a few common threads that different groups of commenters tended to apply to all of the formulas. Among recipients, many comments focused on the lack of data about non-DBE contractors, especially subcontractors. Recipients often noted that they would not have the information needed for the denominator of any of the formulas (i.e. the total number of available businesses). Non-DBE contractors--and industry groups representing them--generally believed that there should be a capacity measure built into any goal setting mechanism. Finally, DBEs--and their industry associations--were concerned that all of the formulas would create goals based only on the current number of DBEs, locking in the effects of past discrimination by ignoring the fact that the lack of opportunities in the past has suppressed the number of DBE firms available today.
Under the proposed rule's Alternative 1, recipients would calculate the percentage of DBE firms in their directories among all firms available to work on their DOT-assisted contracts. Under Alternative 2, recipients would calculate the percentage of all minority-and women- owned firms in certain SIC codes in their areas among all firms in these SIC codes in the same areas. Under Alternative 3, recipients would calculate a percentage based on the average number of DBE firms that had worked on their DOT-assisted contracts in recent years divided by the average number of all firms that had worked on their DOT- assisted contracts in the same period. The SNPRM also proposed that recipients could use other means, such a disparity studies or goals developed by other recipients serving the same area, as a basis for their goals.
Each of the three proposed alternatives received some support, though this was often the rather tepid endorsement of commenters who felt that one or another alternative was the best of a bad lot. Non-DBE contractors often claimed that the alternatives would unfairly increase goals, while DBE contractors often claimed that the same proposals would unfairly decrease goals.
Commenters said that data for determining the denominators of the equations in Alternatives 1 and 2, as well as the numerator in Alternative 2, did not exist and that it would be a major, time- consuming job to begin to obtain the data. Adaptation of existing information from other sources (e.g., Census data) was said to have significant statistical difficulties. The difficulty of getting data on out-of-state firms was emphasized in some comments.
Commenters looked on the alternatives as cumbersome, creating unreasonable administrative burdens, and as producing statistical results that were skewed in various ways. The use of DBE directories as the source of the numerator in Alternative 1 was criticized on the basis that directories may contain firms that never actually participate in DOT-assisted contracts. It was suggested that the number of firms bidding rather than the number of firms certified would be a more reliable guide, but it was also pointed out that, because subcontractors seldom formally bid for work, this data would be hard to obtain. Some commenters proposed adding overall population statistics to the mix.
A significant number of commenters--primarily non-DBE contractors, but including some recipients and other commenters as well--emphasized the need to take ``capacity'' into account. Most popular among these comments was using a capacity version of Alternative 3. These comments did not propose a method of determining the capacity of the firms contracting with the recipient.
In view of the complexity and importance of the goal setting process and the many issues raised by commenters, the Department has decided to adopt a two step process for goal setting. The process is intended to provide the maximum flexibility for recipients while ensuring that goals are based on the availability of ready, willing and able DBEs in each recipient's relevant market. The Department believes that this approach is critical to meeting our constitutional obligation to ensure that the program is narrowly tailored to remedy the effects of discrimination. The first step of the process will be to create a baseline figure for the relative availability of ready, willing and able DBEs in each recipient's market. The second step will be to make adjustments from the base figure, relying on an examination of additional evidence, past experience, local expertise and anticipated changes in DOT-assisted contracting over the coming year.
Step 1: Determining a Base Figure for the Overall Goal
The base figure is intended to be a measurement of the current percentage of ready, willing and able businesses that are DBEs. Ensuring that this figure is based on demonstrable evidence of each recipient's relevant market conditions will help to ensure that the program remains narrowly tailored. To be explicit, recipients cannot simply use the 10 percent national goal, their goal from the previous year, or their DBE participation level from the previous year as their base figure. Instead, all recipients must take an actual measurement of their marketplace, using the best evidence they have available, and derive a base figure that is as fair and accurate a representation as possible of the percentage of available businesses that are DBEs.
There are many different ways to measure the contracting market and assess the relative availability of DBEs. As discussed above, the SNPRM proposed three alternate formulas to measure relative availability, none of which were particularly popular with commenters. In this final rule, the Department is placing primary emphasis on the principles underlying the measurement, mandating only that a measurement of the relative availability of DBEs be made on the basis of demonstrable evidence of relevant market conditions, rather than requiring that any particular procedure or formula be used. The final rule contains a number of examples of how to create a base figure which recipients are free to adopt in their entirety or to use as guidelines for how to devise their own measurement.
There are several reasons we have taken this approach. First, the Department is aware of the differences in available data in various markets across the nation. The flexibility inherent in this approach will ensure that all recipients can use the procedure to set a reasonable goal and allow each recipient to use the best data available to it. As discussed in another section, this rule will also provide for the development of more standard data for future goal setting. Second, for many recipients, setting goals in this way will be a new exercise. By fixing only the basic principle, but allowing the methodology to change, recipients will have the opportunity to fine tune the process each year as their experience grows and the data available to them improve. Finally, the rule makes sure that every recipient will have at least one reasonable and practical goal setting method available to them.
The first example for setting a base figure relies on data sources that are immediately available to all recipients: their DBE directories, and a Census Bureau database that DOT and the Census Bureau will make available to all recipients that wish to use it. This example has its roots in the first two goal setting formulas proposed in the SNPRM. Recipients would first assess the number of ready, willing and able DBEs based on their own directories. For some recipients this will be as simple as counting the number of firms in their directory. For others, particularly those using directories maintained by other agencies, the directories will have to be ``filtered'' for firms involved in transportation contracting. The resulting number of DBEs would become the numerator. The denominator would then be derived from the Census Bureau's County Business Pattern (CBP) database. We will provide user-friendly electronic access to the database via the internet to allow recipients to input the geographic area and SIC codes in which they contract and receive a number for the availability of all businesses.
There are several issues that must be addressed when comparing numbers derived from two different data sources, some of which were raised in the comments on the SNPRM. Recipients will need to ensure that the scope of businesses included in the numerator is as close as possible to the scope included in the denominator. Using as close as possible to the same SIC codes and geographic base is very important. A recipient using its own DBE directory, particularly one that contains only firms in the fields in which it contracts, will still need to determine what fields it will use for the denominator when sorting through the CBP database. The best way to do this would be to examine their contracting program and determine the SIC codes in which they let the substantial majority of their contracts and subcontracts. The geographic area used for both the numerator and the denominator should cover the area from which the recipient draws the substantial majority of its contractors. While it may be sufficient for some state recipients to use their state borders as their contracting area, local transit and airport recipients will rarely have such an obvious choice. Those recipients will need to more carefully examine the geographic area from which they draw contractors and base their calculation of both the numerator and denominator of the equation on the same area.
The Department and the Census Bureau will make the CBP data available in a format that gives recipients as much flexibility as possible to tailor the data to their contracting programs. Recipients will be able to extract the data in one block for all of the SIC codes they expect to contract in, or by individual SIC codes, allowing them to weight the relative availability of DBEs in various fields, giving more weight to the fields in which they spend more money. For example, let us assume a recipient estimates that it will expend 10% of its federal aid funds within SIC code 15, 40% in SIC code 16, 25% in SIC code 17, and the remaining 25% on contracting spread over SIC codes 07, 42 and 87. The recipient could separately determine the relative availability of DBEs for each of the three major construction SIC codes (i.e., 15, 16 and 17) and the relative availability of DBEs in the other three SIC codes grouped together and weight each according to the amount of money to be spent in each area. In this example, the recipient could calculate its weighted base figure by first determining the number of DBEs in its directory for each of the groups, then extracting the availability of CBP businesses for the same groups. It would then perform the following calculation to arrive at a base figure for step one of the goal setting process:
Base=[.10(DBEs in SIC 15) + .40(DBEs in 16) + .25(DBEs in 17) +.25(DBEs in 07,42,87) ]x 100
Figure CBPs in SIC 15 CBPs in 16 CBPs in 17 CBPs in 07,42,87
As has been stated generally, this formula is offered only as an example of a way that a recipient could choose to use the CBP database. Recipients using the CBP data should choose whether to weight their calculation, and whether to do so by individual SIC codes or by groups of SIC codes, based on their own assessment of what method will best fit their spending pattern.¹
¹While it is not statistically necessary to account for 100% of program dollars when performing this type of weighting, the greater the percentage accounted for, the more accurate the resulting calculation will be.
Finally, there is still the question of the propriety of comparing data from two sources as different as DBE directories and the CBP. As mentioned above, some commenters asserted that the directories may contain firms that do not normally perform DOT-assisted contracts. This problem is greatest, of course, for directories maintained by other agencies for purposes beyond DOT-assisted contracting. We believe that the recipient's knowledge of its contracting needs and the contents of its DBE directory will allow it to solve this problem by sorting the directories by SIC code to extract only the firms likely to be interested in DOT-assisted contracting. Any remaining effect from DBEs that are certified in the relevant SIC codes but still do not intend to compete for DOT-assisted contracts will be more than offset by the hurdles involved in actually becoming a DBE. It is important to note here that the certification process itself, with its paperwork, review and on-site inspection, create a filter on the number of existing firms that will be counted in the numerator without there being any equivalent filter culling firms out of the denominator. Ultimately, the Department chose these two data sources for the example because; while they may not be perfect, they represent the best universally available current data on both the presence of DBEs and the presence of all businesses in local markets. Any recipient that believes it has available to it better sources of local data from which to make a similar calculation for its base figure is encouraged to use them.
The second example for calculating a base figure is using a bidders list to determine the relative availability of DBEs. The concept is similar to the one described above. The recipient would divide the number of available ready, willing and able DBEs by the number for all firms. The difference is that instead of measuring availability by DBE certifications and Census data, the recipient would measure availability by the number of firms that have directly participated in, or attempted to participate in, DOT-assisted contracting in the recent past. This approach has its roots in Alternative 3 from the SNPRM. Of fundamental importance to this approach is that the recipient would need to include all firms that have sought DOT-assisted contracts, regardless of whether they did so by bidding on a prime contract or quoting a job as a subcontractor. Because most DOT recipients derive the substantial majority of their DBE participation through subcontracting, it is absolutely essential that all DBE and non-DBE firms that quote subcontracts be included in the bidders list.² Bidders lists are a very focussed measure of ready, willing and able firms because they filter the pool of available firms by requiring a demonstration of their ability to participate in the process through tracking and identifying contracting opportunities, understanding the requirements of a particular job and assembling a bid for it. Another attractive feature of the bidding ``filter'' is that it applies equally to both DBEs and non-DBEs.
²To prevent any confusion, it is important to note that the DBE program does not use the so-called ``benchmarking'' system employed in direct Federal procurement. The benchmarking system relies on a unique database created specifically for use in the federal procurement program.
The third example included in the final rule for setting a base figure is using data derived from a disparity study. As was discussed in the SNPRM, the Department is not requiring recipients to do a disparity study, but is only making clear that use of disparity study data by recipients that have them or choose to conduct them is a valid means of setting a goal. Disparity studies generally contain a wide array of statistical data, as well as anecdotal data and analysis that can be particularly useful in the goal setting process. We list disparity studies here, not because they are needed to justify operating the DBE program--Congress has already established the compelling need for the DBE program--but because the data a good disparity study provides can be an excellent guide for a recipient to use to set a narrowly tailored goal.
The Department will not set out specific requirements for what data or analysis is required before a disparity study can be used for setting a goal, because we believe that the design and conduct of the study is best left to the local officials and the professional organizations with which they contract to conduct the studies. Instead, we again offer simple general principles that should apply to all studies used for goal setting. Any study data relied on in the goal setting process should be as recent as possible and be focussed on the transportation contracting industry. When setting the goal, first use the study's statistical evidence to set a base figure for the relative availability of DBEs. Other study information, whether it is anecdotal data, analysis or statistical information about related fields, should be included when making adjustments to the base figure (discussed in more detail below), but not included in the base figure for the relative availability of DBEs.
The last specific example included in the rule is using the goal of another recipient as the base figure for goal setting. This option was also included in the SNPRM. It is intended to avoid duplicative work and to lighten the burden the goal setting process might put on smaller recipients. It is important to note that a recipient could only use another recipient's goal if it was set in accordance with this rule and the other recipient performed similar contracting in a similar market area. Using another recipient's approved goal would only satisfy the first step of the goal setting process. It would serve as the base figure, and could not be used to skip over step two of the process. The recipient would need to examine the same additional evidence it would otherwise use to determine whether to adjust its goal from the base figure, as well as being required to make adjustments to account for differences in its local market or contracting program.
The final rule also maintains the option of devising an alternative method of calculating a base figure for the goal setting process. Explicitly listing this option serves to emphasize the point that the options in the rule are examples meant as guidelines intended to ensure maximum flexibility for recipients. Recipients can use this option to take advantage of their unique expertise or any unique source of data that they have that may not be available to other recipients. The concerned operating administration will review and approve the proposals of recipients that believe they can calculate a base figure that will better reflect their relevant market than any of the examples provided in this rule. Approval will be contingent on the proposals following the same principles that apply to any recipient: the methodology must be based on demonstrable data of relevant market conditions and be designed to reach a goal that the recipient would expect DBEs to achieve in the absence of discrimination.
Step 2: Adjusting the Base Figure
As alluded to above, measuring the relative availability of DBEs to derive a base figure is only the first step of the goal setting process. To ensure that they arrive at goals that truly and accurately reflect the participation they would expect absent the effects of discrimination, recipients must go beyond the formulaic measurement of current availability to account for other evidence of conditions affecting DBEs. To accomplish this second step, recipients must first survey their jurisdiction to determine what types of relevant evidence is available to them. Then, relying on their own knowledge of their contracting markets they must review the evidence to determine whether either an up or down adjustment from the base figure is needed.
One universally available form of evidence that all recipients should consider is the proven capacity of DBEs to perform work on DOT- assisted contracts. All recipients have been tracking and reporting the dollar volume of work that is contracted and subcontracted to DBEs each year. Viewed in isolation, the past achievements of DBEs do not reflect the availability of DBEs relative to all available businesses, but it is an important and current measure of the ability of DBEs to perform on DOT-assisted contracts.
Though not universally available, there are hundreds of existing disparity studies that contain a wealth of statistical and anecdotal evidence on the utilization of disadvantaged businesses. In addition to being a possible source of data for Step 1 of the goal setting process, disparity studies should be considered during Step 2 of the process. The base figure from Step 1 is intended to determine the relative availability of DBEs. The data and analysis in a disparity study can help a recipient determine whether those existing businesses are under- or over-utilized. If a recipient has a study with disparity ratios showing that existing DBEs are receiving significantly less work than expected, an upward adjustment from the base figure is called for. Similarly, if the disparity ratio shows overutilization, a downward adjustment to the base figure would be warranted. The anecdotal evidence and analysis of contracting requirements and conditions that may have a discriminatory impact on DBEs are also important sources that should be examined when determining what adjustment to make to the base figure.³ Finally, disparity studies that are conducted within a recipient's jurisdiction should be examined even if they were not done specifically for the recipient. For example, a state highway agency may find useful data and analysis in either a statewide disparity study covering other agencies or in a disparity study examining contracting in a county or city within the state.
³ It is important to note that adjusting the goal is only part of the response a recipient should make to evidence of discriminatory barriers for DBEs. All recipients have a primary responsibility to ensure non-discrimination in their programs and should act aggressively to remove any discriminatory barriers in their programs.
If a recipient uses another recipient's goal as its base figure under Step 1 of the goal setting process, it will have to make additional adjustments to ensure that its final goal is narrowly tailored to its market and contracting program. For example, if a local transit or airport authority adopts a statewide goal as its base figure, it must determine the extent that local relative availability of DBEs differs from the relative availability of DBEs in the contracting area relied on by the state. The local recipient would also need to examine the differences in the type of contracting work in its program and determine whether there are significant differences in the relative availability of DBEs in any fields that are unique to its program--or unique to the program of the other recipient. Similarly, if one local recipient used the goal of another local recipient in the same market as its base figure, it would also need to adjust for differences in the contracting fields used by the two programs.
Finally, the rule contains a brief list of other types of data a recipient could consider when adjusting its base figure to arrive at an overall goal. The list is by no means intended to be exhaustive. Instead, it is meant as a guide to the types of information a recipient should look for in Step 2 of the goal setting process. There is a wide array of relevant local, regional and national information about the utilization of disadvantaged businesses. Recipients are encouraged to cast as wide a net as they can to carefully examine their contracting programs and the public and private markets in which they operate.
Additional Goal Setting Issues
The Department proposed, in both the 1992 NPRM and the 1997 SNPRM, that overall goals be calculated as a percentage of DOT funds a recipient expects to expend in DOT-assisted contracts. This is different from the existing part 23 rule, which asked recipients to set overall goals on the basis of all funds, including state and local funds, to be expended in DOT-assisted contracts. This change is for accounting and administrative convenience and is not intended to have a substantive effect on the program. While not the subject of many comments, those who did comment on the proposal favored the change. The final rule adopts this approach.
A few recipients commented that public participation concerning goal setting was bothersome. Nevertheless, we view it as an essential part of the goal setting process. There are many stakeholders involved in setting goals, and it is reasonable that they should be involved in the process and have an opportunity for comment. The part 23 provision requiring getting a state governor's approval of a goal of less than 10 percent has been eliminated, both because overall goals are no longer tied to the national 10 percent goal and to reduce administrative burdens.
The goal setting provision of the final rule continues to direct recipients to set one annual overall goal for DBEs, rather than group- specific goals separating minority and women-owned businesses.