NERA Economic Consulting National Economic Research Associates, Inc. th 1006 East 39 St. Austin, TX 78751 512 371 8995 Fax 512 371 8996 www.nera.com RACE, SEX, AND BUSINESS ENTERPRISE: EVIDENCE FROM THE STATE OF WASHINGTON Prepared for the Washington State Department of Transportation by NERA Economic Consulting FINAL REPORT--October 20, 2005 NERA Economic Consulting TABLE OF CONTENTS LIST OF TABLES.......................................................................................................................................................... II ACKNOWLEDGEMENTS .......................................................................................................................................... V I. II. A. B. C. III. A. B. C. IV. V. INTRODUCTION.................................................................................................................................................... 1 DEFINING THE RELEVANT MARKETS ...................................................................................................... 5 PREPARING THE MASTER CONTRACT/SUBCONTRACT DATABASE ....................................................................... 5 PRODUCT MARKET DEFINITION ........................................................................................................................... 7 GEOGRAPHIC MARKET DEFINITION ..................................................................................................................... 8 IDENTIFYING BUSINESSES IN THE RELEVANT MARKETS ............................................................. 9 ESTIMATE THE TOTAL NUMBER OF BUSINESSES IN THE MARKET ..................................................................... 9 IDENTIFY LISTED DBES ..................................................................................................................................... 10 VERIFY LISTED DBES AND ESTIMATE UNLISTED DBES .................................................................................. 12 ESTIMATING BASELINE DBE AVAILABILITY ..................................................................................... 16 DBE PARTICIPATION IN WSDOT CONTRACTING AND SUBCONTRACTING, FFY 1999-2003 18 DISPARITIES IN MBE BUSINESS FORMATION AND BUSINESS OWNER EARNINGS ............. 19 REVIEW OF RELEVANT LITERATURE ................................................................................................................. 20 RACE AND SEX DISPARITIES IN EARNINGS ........................................................................................................ 22 RACE AND SEX DISPARITIES IN BUSINESS FORMATION .................................................................................... 30 ESTIMATES OF ADJUSTED DBE AVAILABILITY ................................................................................................ 37 TABLES .............................................................................................................................................................. 38 CONCLUSION ................................................................................................................................................ 98 REFERENCES .................................................................................................................................................. 100 VI. A. B. C. D. VII. VIII. IX. NERA Economic Consulting LIST OF TABLES TABLE 1. PRODUCT MARKET FOR ALL WSDOT CONTRACTS ...................................................................................... 39 TABLE 2. PRODUCT MARKET FOR WSDOT CONSTRUCTION CONTRACTS ................................................................... 40 TABLE 3. PRODUCT MARKET FOR WSDOT CONSULTING CONTRACTS........................................................................ 41 TABLE 4. DISTRIBUTION OF WSDOT CONTRACT DOLLARS BY CATEGORY ................................................................ 42 TABLE 5. COUNTY DISTRIBUTION OF WSDOT CONTRACT DOLLARS .......................................................................... 43 TABLE 6. TOTAL BUSINESSES AND INDUSTRY WEIGHT, BY SIC CODE ........................................................................ 45 TABLE 7. CONSTRUCTION BUSINESSES AND INDUSTRY W EIGHT, BY SIC CODE.......................................................... 46 TABLE 8. CONSULTING BUSINESSES AND INDUSTRY WEIGHT, BY SIC CODE .............................................................. 47 TABLE 9. LISTED DBES AND INDUSTRY W EIGHT, BY SIC CODE .................................................................................. 48 TABLE 10. LISTED CONSTRUCTION DBES & INDUSTRY WEIGHT, BY SIC CODE ........................................................ 49 TABLE 11. LISTED CONSULTING DBES & INDUSTRY W EIGHT, BY SIC CODE ............................................................. 50 TABLE 12. LISTED DBE PERCENTAGE & INDUSTRY WEIGHT, BY SIC CODE .............................................................. 51 TABLE 13. LISTED CONSTRUCTION DBE PERCENTAGE & INDUSTRY WEIGHT, BY SIC CODE ................................... 52 TABLE 14. LISTED CONSULTING DBE PERCENTAGE & INDUSTRY W EIGHT, BY SIC CODE........................................ 53 TABLE 15. LISTED DBE SURVEY--A MOUNT OF MISCLASSIFICATION, BY SIC CODE G ROUPING .............................. 54 TABLE 16. LISTED DBE SURVEY--A MOUNT OF MISCLASSIFICATION, BY H IGHWAY REGION ................................... 55 TABLE 17. LISTED DBE SURVEY--A MOUNT OF MISCLASSIFICATION, BY PUTATIVE DBE TYPE .............................. 56 TABLE 18. UNCLASSIFIED BUSINESSES SURVEY --BY SIC CODE GROUPING ............................................................. 57 TABLE 19. UNCLASSIFIED BUSINESSES SURVEY --BY HIGHWAY REGION .................................................................. 58 TABLE 20. UNCLASSIFIED BUSINESSES SURVEY--BY RACE AND SEX ......................................................................... 59 TABLE 21. CALCULATION SUMMARY--OVERALL ......................................................................................................... 60 TABLE 22. CALCULATION SUMMARY--CONSTRUCTION ............................................................................................... 61 NERA Economic Consulting TABLE 23. CALCULATION SUMMARY--CONSULTING ................................................................................................... 62 TABLE 24. ESTIMATED DBE AVAILABILITY FOR WSDOT ........................................................................................... 63 TABLE 25. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, G ROSS CONTRACT AMOUNT...................................................................................... 64 TABLE 26. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, NON-SUBCONTRACTED DOLLAR A MOUNTS ............................................................. 65 TABLE 27. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--FEDERALLY-FUNDED ONLY, PRIME AND SUBCONTRACTS, FIRST-TIER ONLY ................................................................................................... 66 TABLE 28. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--N ON-FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, GROSS CONTRACT A MOUNT.......................................................................... 67 TABLE 29. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--N ON-FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, NON-SUBCONTRACTED DOLLAR A MOUNTS ................................................. 68 TABLE 30. ESTIMATED DBE U TILIZATION ON WSDOT CONSTRUCTION PROJECTS--N ON-FEDERALLY-FUNDED ONLY, PRIME AND SUBCONTRACTS, FIRST-TIER ONLY ....................................................................................... 69 TABLE 31. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, G ROSS CONTRACT AMOUNT...................................................................................... 70 TABLE 32. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, NON-SUBCONTRACTED DOLLAR A MOUNTS ............................................................. 71 TABLE 33. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--FEDERALLY-FUNDED ONLY, PRIME AND SUBCONTRACTS, FIRST-TIER ONLY ................................................................................................... 72 TABLE 34. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--NON-FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, G ROSS CONTRACT AMOUNT...................................................................................... 73 TABLE 35. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--NON-FEDERALLY-FUNDED ONLY, PRIME CONTRACTS ONLY, NON-SUBCONTRACTED DOLLAR A MOUNTS ............................................................. 74 TABLE 36. ESTIMATED DBE U TILIZATION ON WSDOT CONSULTING PROJECTS--NON-FEDERALLY-FUNDED ONLY, PRIME AND SUBCONTRACTS, FIRST-TIER ONLY ................................................................................................... 75 TABLE 37. ANNUAL WAGE EARNINGS REGRESSIONS, ALL INDUSTRIES, 2000............................................................ 76 TABLE 38. ANNUAL WAGE EARNINGS REGRESSIONS, ALL INDUSTRIES, 1979-1991 .................................................. 77 TABLE 39. ANNUAL WAGE EARNINGS REGRESSIONS, ALL INDUSTRIES, 1992-2002 .................................................. 78 TABLE 40. ANNUAL WAGE EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 2000 ................ 79 TABLE 41. ANNUAL WAGE EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1979-1991 ...... 80 NERA Economic Consulting TABLE 42. ANNUAL WAGE EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1992-2002 ...... 81 TABLE 43. ANNUAL BUSINESS OWNER EARNINGS REGRESSIONS, ALL INDUSTRIES, 2000......................................... 82 TABLE 44. ANNUAL BUSINESS OWNER EARNINGS REGRESSIONS, ALL INDUSTRIES, 1979-1991............................... 83 TABLE 45. ANNUAL BUSINESS OWNER EARNINGS REGRESSIONS, ALL INDUSTRIES, 1992-2002............................... 84 TABLE 46. BUSINESS OWNER EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 2000 ............ 85 TABLE 47. BUSINESS OWNER EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1979-1991... 86 TABLE 48. BUSINESS OWNER EARNINGS REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1992-2002... 87 TABLE 49. SELF-EMPLOYMENT RATES IN 2000 FOR SELECTED RACE AND SEX GROUPS: A LL INDUSTRIES; UNITED STATES AND THE STATE OF WASHINGTON ............................................................................................................ 88 TABLE 50. SELF-EMPLOYMENT RATES IN 2000 FOR SELECTED RACE AND SEX GROUPS: CONSTRUCTION AND RELATED INDUSTRIES; UNITED STATES AND THE STATE OF W ASHINGTON ........................................................ 89 TABLE 51. BUSINESS FORMATION REGRESSIONS, A LL INDUSTRIES, 2000 ................................................................... 90 TABLE 52. BUSINESS FORMATION REGRESSIONS, A LL INDUSTRIES, 1979-1991 ......................................................... 91 TABLE 53. BUSINESS FORMATION REGRESSIONS, A LL INDUSTRIES, 1992-2002 ......................................................... 92 TABLE 54. BUSINESS FORMATION REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 2000 ....................... 93 TABLE 55. BUSINESS FORMATION REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1979-1991 ............. 94 TABLE 56. BUSINESS FORMATION REGRESSIONS, CONSTRUCTION AND RELATED INDUSTRIES, 1992-2002 ............. 95 TABLE 57. A CTUAL AND POTENTIAL BUSINESS FORMATION RATES--WASHINGTON CONSTRUCTION AND CONSULTING MARKETS .......................................................................................................................................... 96 TABLE 58. COMPARISON OF BASELINE TO A DJUSTED DBE A VAILABILITY FOR WSDOT .......................................... 97 NERA Economic Consulting ACKNOWLEDGEMENTS The work of our study team was made possible through the assistance of the following persons: Brenda Nnambi, Jim Medina, John Anderson, Pete Anthony, Mike Rice, and their respective co-workers and staff at the Washington State Department of Transportation; Madelon Barton at the Washington State Department of Revenue, and Attorney Colette Holt of Colette Holt and Associates in Chicago. We would also like to acknowledge our own study team--NERA Special Consultant Dr. David Blanchflower at Dartmouth, NERA Research Analysts Debbie Norris and Kim Stewart in Austin, and SRBI Project Manager Andrew Evans in Fort Myers. Their contributions were instrumental to the success of this study. Jon Wainwright, Vice President, NERA NERA Economic Consulting I. INTRODUCTION The Washington State Department of Transportation (WSDOT) commissioned NERA Economic Consulting to perform this study in compliance with United States Department of Transportation (USDOT) regulations. WSDOT is charged with providing a safe, efficient, and effective statewide transportation system, and as such is responsible for the planning, construction, and maintenance of an extensive transportation network throughout Washington State. This network includes over 7,000 miles of highways and roads, 34 tunnels, 43 rest areas, 97,500 acres of roadside lands and associated drainage structures, 10 ferry routes, 20 ferry terminals, one ferry repair facility, and 16 emergency airports.1 Between 2005 and 2011, WSDOT expects to spend almost $2.6 billion for the highway improvement, highway preservation, and ferry construction projects necessary to carry out its transportation mission.2 Each federal fiscal year, the Federal Highway Administration (FHWA) and the other modal agencies of USDOT provide significant levels of federal funding to WSDOT to support its construction and preservation activities. Between FFY 2002 and FFY 2004, for example, WSDOT received almost $1.8 billion from the FHWA. As a recipient of such funds, WSDOT is required to comply with the regulations pertaining to the USDOT's Disadvantaged Business Enterprise (DBE) Program. The primary concern of the DBE Program is to create a level playing field for the utilization of businesses owned by socially and economically disadvantaged persons, including members of certain minority groups and women, on contracts that are funded in part or in whole by USDOT. In 1999, USDOT adopted a comprehensive revision of the DBE Program.3 WSDOT must set an overall, annual aspirational percentage goal for DBE participation on its USDOT-assisted contracts that are narrowly tailored to WSDOT's particular circumstances 1 Washington State Department of Transportation and Washington State Transportation Commission. February 2002. Washington's Transportation Plan, 2003-2022. pp. 1, 11-12. Washington State Department of Transportation. February 2005. Measures, Markers and Mileposts: The Gray Notebook for the Quarter Ending December 31, 2004. p. 22. 49 Code of Federal Regulations (CFR), part 26. 2 3 1 NERA Economic Consulting and based on demonstrable evidence of availability--i.e. the percentage of relevant businesses owned by minorities and/or women in WSDOT's geographic market area.4 The process for determining availability is twofold.5 First, WSDOT must make a determination of the baseline percentage of firms in its relevant market area that are or could become certified as DBEs. Second, WSDOT must consider other relevant information and make a determination about whether, and if so by how much, the baseline figure should be adjusted upward or downward in order to set an overall goal that is consistent with what would be expected in a market that is race- and sex- neutral, i.e., DBE availability "but for" discrimination. 6 This two-step method requires WSDOT to set a DBE goal that prevents under-utilization of DBEs and over-utilization of DBEs to the exclusion of non-DBEs. Under the regulations, if an agency exceeds its overall goal for two consecutive years through the use of contract-specific DBE participation goals, it must proportionately reduce its use of contract-specific goals in the following year.7 For this study, NERA used minority-owned and women-owned business (MWBE) availability as a proxy for DBE availability. The MWBE and DBE populations have a very high degree of correlation and overlap. There are two differences worth noting, however. First, to be certified as a DBE under Part 26 a business owner's personal net worth, excluding equity in the owner's primary residence and in the business seeking certification, cannot exceed $750,000.8 Hence, not all MWBEs are eligible for certification as DBEs. In practice, however, very few households--especially minority households--have net worth levels in excess of $750,000, especially when home equity and business equity are excluded from the measure.9 Second, it is possible for businesses owned by non-minority males, such 4 5 6 7 8 9 49 CFR ? 26.45. Id. Ibid. 49 C.F.R. ? 26.51(f). 49 CFR ? 26.67. According to the Federal Reserve's 1993 National Survey of Small Business Finances, about 6 percent of White-male-owned small businesses, 2.6 percent of White-female-owned small businesses, and 3 percent of non-White-owned small businesses have business equity in excess of $750,000. Further, Census Bureau data show that the median net worth of Black and Hispanic households is much less (continued...) 2 NERA Economic Consulting as businesses owned by disabled persons, to become certified DBEs if they can establish that they meet the regulatory criteria to be considered socially and economically disadvantaged.10 Hence, not all DBEs are necessarily MWBEs. In practice, however, since so few MWBEs have net worth levels in excess of $750,000 and a substantial number of businesses owned by socially and economically disadvantaged non-minority males could potentially seek DBE certification NERA's method may understate DBE availability to a small degree.11 NERA's approach to availability measurement reflects USDOT's compliance advice. According to the USDOT's guidance, "... if you have data about the number of minority and women-owned businesses (regardless of whether they are certified as DBEs) in your market area, or DBEs in your market area that are in other recipients' Directories but not yours, you can supplement your Directory data with this information. Doing so may provide a more complete picture of the availability of firms to work on your contracts than the data in your Directory alone."12 The remainder of this report is organized as follows. Section II describes the assembly of the contract and subcontracting database and how the definition of the relevant (...continued) than the median for White households. Very few Black or Hispanic households have net worth above even $500,000. Only 0.2 percent of Black households and 0.5 percent of Hispanic households have a net worth greater than $500,000--compared to a figure of 4 percent for White households. Overall, the median net worth for White households is approximately seven times higher than that of Black or Hispanic households. (See U.S. Census Bureau, "Percent Distribution of Household Net Worth, by Amount of Net Worth and Selected Characteristics: 1995," INTERNET: http://www.census.gov/hhes/www/wealth/1995/wlth95-4.html and U.S. Census Bureau, "Median Value of Assets for Households, by Type of Asset Owned and Selected Characteristics: 1995," INTERNET: http://www.census.gov/hhes/www/wealth/1995/wlth95-1.html). More recent Federal Reserve Board data also document that the net worth of White households is much greater than that of Black or Hispanic households. The Federal Reserve's 1998 Survey of Consumer Finances found that the median net worth of non-minority households was $94,900 and the mean net worth was $334,400. For minority households, the median net worth was $16,400 and the mean net worth was $101,700 (See Kennickell, Arthur B., Starr-McCluer, Martha, and Surette, Brian J., "Recent Changes in U.S. Family Finances: Results from the 1998 Survey of Consumer Finances," Federal Reserve Bulletin, January 2000). 10 11 12 49 CFR ? 26.67 and Appendix E. For ease of exposition, we shall use the term DBE throughout the remainder of the report. See INTERNET: http://osdbu.dot.gov/business/dbe/hottips.cfm (emphasis added). This information was released as official guidance by USDOT. See 49 CFR ?26.9. 3 NERA Economic Consulting markets. Section III describes the methods employed to estimate baseline DBE availability and Section IV presents a summary of these methods and the principal results of the availability analysis (step 1). Section V describes the compelling interest evidence considered concerning a possible Step 2 adjustment of the baseline availability figures. At WSDOT's request, we report estimates of DBE availability for contract, subcontract, and supplier opportunities in construction and architectural/engineering design and other professional construction-related consulting. 4 NERA Economic Consulting II. DEFINING THE RELEVANT MARKETS The first step in estimating DBE availability is to define the relevant markets for WSDOT's federally-assisted contracting. Markets have a product and a geographic dimension, both of which were considered in constructing our estimates of DBE availability.13 Once the appropriate markets have been defined, we can estimate the number of businesses present in those markets as well as the number that are owned by minorities or women. Finally, WSDOT contract expenditure data are used to develop dollar-based weights for each relevant industry and county. These weights are combined and then used to calculate overall weighted average DBE availability for the State of Washington and each of its six highway regions. A. Preparing the master contract/subcontract database In order to identify the product and geographic markets relevant to WSDOT, we assembled a master database of WSDOT's contracting and subcontracting activity awarded between Federal Fiscal Year (FFY) 1999 and FFY 2003. This section describes the types of federally-assisted WSDOT projects that are included in this master contract/subcontract database: (1) Construction and (2) Architectural/Engineering Design and Other Construction-Related Consulting Services. We use FFY98-FFY03 data from both categories to identify the industries in WSDOT's product market and the counties in its geographic market. 1. Construction NERA worked with WSDOT construction staff to identify all federally-assisted construction contracts awarded from October 1998 through September 2003 and extracted a profile on each of them from WSDOT's Construction Contract Information System (CCIS). A total of 624 such construction contracts were awarded during that period with a value of more than $1.52 billion. For each contract, our profile included the unique contract 13 See, for example, Areeda, Phillip, and Louis Kaplow, Antitrust Analysis: Problems, Text, Cases, Boston: Little, Brown and Company, 4th Edition, 1988. 5 NERA Economic Consulting identification number, unique business identification number, business name, business address, award date, contract award amount and federal assistance participation percentage. WSDOT has wisely also collected and retained information on the first-tier subcontractors and suppliers for each CCIS contract, including their unique business identification number, business name, business address and contract award amount. In most instances, the CCIS file also indicated each firm's DBE status, including race and sex. Next, we cross-referenced the businesses in the CCIS file with the State Business Records Database--a file of all active businesses registered with the Department of Revenue--in order to obtain a primary Standard Industrial Classification (SIC) code for each firm.14 SIC codes for the relatively small number of firms that could not be matched in this manner were identified through manual lookups in Dun & Bradstreet and ABI-Inform. 2. Architectural/Engineering Design and Other Construction-Related Consulting Services We also worked with WSDOT Consultant Services Section (CSS) staff to identify all federally-assisted contracts for architectural/engineering design and related professional consulting services (hereafter, "Consulting") awarded between October 1998 and September 2003. We obtained data for 89 such contracts executed during that period with an aggregate value of more than $107 million. As with the construction contracts, we received data including the unique contract identifier, unique business identifier, business name, business address, contract approval date, contract award amount and federal funds participation percentage. As with construction projects, WSDOT has wisely collected and retained first-tier subcontractor and supplier data for consulting projects. The first-tier sub-consultant data we obtained included the unique business identifier, business name, business address, contract award date, contract award amount and DBE status. 14 We assigned or confirmed each firm's type of work using four-digit SIC codes, which are the most detailed level available in the SIC system. 6 NERA Economic Consulting Next, we assigned and/or confirmed SIC codes for each consultant and subconsultant in this database, using the sources identified above as well as descriptions in the CSS data concerning the type of work being performed. B. Product Market Definition Based on the SIC codes assigned to each contractor and subcontractor in the master database, we estimated product market weights according to the proportion of total contract and subcontract dollars attributable to each SIC code. These weights show the relative importance, in dollars, of the activity represented in each SIC code. In Construction, we identified 97 distinct SIC codes within the 624 contracts we studied. Of these 97 SIC codes, however, 26 account for 99 percent of the total dollars spent. For this study, we take these 26 SIC codes to represent WSDOT's Construction product market. In Consulting, we identified 21 distinct SIC codes within the 89 contracts we studied. Of these 21 SIC codes, however, 6 account for 99 percent of the total dollars spent. For this study, we take these 6 SIC codes to represent WSDOT's Consulting product market. Although numerous industries play a role in WSDOT's contracting activities, it is clear that contracting opportunities are not distributed evenly among them. The distribution of contract expenditures is, in fact, highly skewed. Overall (Table 1), four industries account for two-thirds of total contract and subcontract spending by WSDOT during the study period. In Construction (Table 2), a single industry--highway and street construction--accounts for almost 42 percent of all contracting expenditures, and the top five industries account for almost 75 percent. Concentration is even more prevalent in Consulting (Table 3), where a single industry-- Engineering Services--accounts for over 93 percent of all contracting expenditures. 7 NERA Economic Consulting C. Geographic Market Definition We turn next to a determination of the geographic dimension of WSDOT's contracting markets. We used the master contract/subcontract database, as described above in Section II.A, to obtain the zip codes for each contractor and subcontractor in the database. We then disaggregated the database by state, highway region, county, and Metropolitan Statistical Area (MSA) and calculated the percentage of WSDOT contract dollars awarded to businesses in different geographic areas. Table 4 presents the results of these calculations. Businesses located in Washington State account for the vast majority of WSDOT's contracting expenditures, regardless of category. As shown in Table 4, WSDOT awarded 93.7 percent of its construction dollars during the study period to contractors with businesses located in Washington.15 For consulting contracts, the figure was 92.4 percent, 16 and the combined figure is 93.6%. Based on these results, we will define WSDOT's geographic market to be the State of Washington for purposes of estimated availability. Within the State of Washington, there is still considerable county-to-county variation in WSDOT's contract spending. Table 5 shows, for example, that businesses located in King, Kitsap, Snohomish and Pierce Counties (greater Seattle) account for relatively more construction contract and subcontract dollars than do businesses located elsewhere in the State, and that consulting work, in particular, is centered strongly around King County.17 15 After Washington, the most important states in terms of contract dollars were Oregon (4.1 percent), Idaho (0.8 percent), California (0.7 percent), and Utah (0.3 percent). After Washington, the most important states in terms of contract dollars were California (5.2 percent), Virginia (1.2 percent), and Illinois (0.5 percent). No contractors or subcontractors were located in the Washington counties of Adams, Ferry, Garfield, or San Juan. 16 17 8 NERA Economic Consulting III. IDENTIFYING BUSINESSES IN THE RELEVANT MARKETS The DBE availability percentage (unweighted) is defined as the number of DBEs divided by the total number of businesses in the counties and industries relevant to WSDOT's contracting activities.18 Determining the total number of businesses in the relevant markets is more straightforward than determining the number of minority- or women-owned businesses in those markets. The latter task has three main parts: (1) identify all listed DBEs in the relevant market; (2) verify the ownership status of listed DBEs; and (3) estimate the number of unlisted DBEs in the relevant market. This section describes, in turn, how both tasks were accomplished. A. Estimate the Total Number of Businesses in the Market We used Dun & Bradstreet's MarketPlace database to determine the total number of businesses operating in the relevant geographic and product markets (these markets were discussed in the previous section). MarketPlace is a comprehensive database of U. S. businesses. This database, which contains over 13 million records, is updated continuously, and Dun & Bradstreet issues a revised version each quarter. For this study, we used data for the second quarter of 2004. Each record in MarketPlace represents a business and includes the company name, address, telephone number, primary four-digit SIC code, secondary SIC code(s) (if any), business type, DUNS Number (a unique number assigned to each business by Dun & Bradstreet) and other descriptive information. Dun & Bradstreet gathers and verifies information from many different sources. These sources include annual management interviews, payment experiences, bank account information, filings for suits, liens, judgments and bankruptcies, news items, the U. S. Postal Service, utility and telephone service, business registrations, corporate charters, Uniform Commercial Code filings, and records of the Small Business Administration and other governmental agencies. We used the MarketPlace database to identify the total number of businesses in each four-digit SIC code to which we had assigned a product market weight.19 Table 6 18 19 To yield a percentage, the resulting figure is multiplied by 100. These weights are described above in Section II.B. 9 NERA Economic Consulting shows the number of businesses identified in each SIC code, along with the associated industry weight (all contracting combined). Comparable data for construction and consulting appear in Tables 7 and 8, respectively. The industry weights that are listed are the same as those appearing above in Tables 1-3, respectively. B. Identify Listed DBEs As extensive as it is, MarketPlace itself does not adequately identify all businesses owned by minorities or women. Although many such businesses are correctly identified in MarketPlace, experience has demonstrated that many more are missed. For this reason, several additional steps were required to identify the appropriate percentage of DBEs in the relevant market. First, NERA completed an intensive regional search for information on minorityowned and woman-owned businesses in Washington State and surrounding areas. Beyond the information already in MarketPlace, NERA collected lists of DBEs from WSDOT as well as other public and private entities in and surrounding the State of Washington. Specifically, directories were included from:20 Washington State Department of Transportation, Washington State Office of Minority & Women's Business Enterprise, Associated General Contractors of Washington, Business Research Services National Directory of Minority-Owned Businesses, Business Research Services National Directory of Women-Owned Businesses, CalTrans, the City Olympia, the City of Portland Sheltered Market Program, the City of Seattle Boost Program, the City of Seattle Vendor & Contractor Registration list, the City of Spokane, the City of Tacoma, the City of Vancouver, the Contractor Development & Competitiveness Center of the Urban League of Metropolitan Seattle, Diversity Information Resources, the federal government's Central Contractor Registration database, the Idaho Transportation Department, King County, the 20 We also obtained information from certain entities that was duplicative of either Dun & Bradstreet or one or more of the other sources listed above. These entities include the City of Olympia, the City of Portland Sheltered Market Program, the City of Spokane, the City of Vancouver, King County, the Kroger Company, Nordstrom's Department Stores, Pepsico, the Port of Portland, the Port of Seattle, the Port of Tacoma, Qwest Communications, Raytheon, Sound Transit Diversity Programs, the Tacoma Housing Authority, Thurston County, the U.S. Army Corps of Engineers, the University of Washington, Washington Mutual, W.W. Grainger Co., and the Xerox Corporation. 10 NERA Economic Consulting Montana Department of Transportation, National Association of Women Business OwnersInland Northwest Chapter, National Association of Women in Construction (Puget Sound, Spokane, Tri-Cities, and Yakima Valley Chapters), the National Center for American Indian Economic Development, the Nevada Department of Transportation, the Northwest Native American Business Development Center, the Oregon Office of Minority, Women and Emerging Small Business, the Port of Portland, the Port of Seattle, the Port of Tacoma, the Seattle Monorail Project, Sound Transit Diversity Programs, the South Puget Sound Hispanic Chamber of Commerce, the Tabor 100 (Northwest Association of AfricanAmerican Businesses), the Tacoma Housing Authority, Thurston County, the U.S. Army Corps of Engineers, the University of Washington, the Washington State Hispanic Chamber of Commerce, Women Business Owners of Puget Sound, and the Women's' Business Enterprise National Council.21 We will refer to the DBE businesses identified in this manner as "listed" DBEs. Tables 9-11 provide the total number of listed DBEs by SIC code--overall, and for construction and consulting, respectively.22 If the listed DBEs identified in the three previous tables are all in fact DBEs and are the only DBEs among all the businesses identified in Tables 6-8, then an estimate of "listed" DBE availability would be calculated as shown in Tables 12-14. The availability figure in these tables is simply the number of listed DBEs (taken from Tables 9-11, respectively) divided by the total number of businesses in the relevant market (taken from 21 A number of organizations we contacted declined to participate in this study or were otherwise unresponsive to our (or WSDOT's) repeated requests. These include: Bank of America Supplier Diversity Program, the Black Chamber of Commerce Pacific Northwest Chapter, the Boise Cascade Corp Supplier Diversity Program., CH2M Hill, Chevron/Texaco Supplier Diversity Program., the City of Bellevue, Coca Cola Enterprises Supplier Diversity Program, the Community Capital Development SMWBE list, Conoco/Phillips Supplier Diversity Program, Georgia-Pacific Supplier Diversity Program, Howard S. Wright Construction Supplier Diversity Program, Microsoft Supplier Diversity Program, the National Association of Minority Contractors, the National Association of Women in Construction, Takoma Chapter, the National Minority Business Council, Nike Supplier Diversity Program, Nordstrom Department Stores Supplier Diversity Program, the Northwest Minority Business Council, the Oregon Association of Minority Entrepreneurs, Safeco Insurance Company Supplier Diversity Program, Seattle Mariners Supplier Diversity Program, the Seattle/Washington State Minority Business Development Center, Starbucks Supplier Diversity Program, the Boeing Company Supplier Diversity Program, W.W. Grainger Co. Supplier Diversity Program, and the Wells Fargo Supplier Diversity Program. The industry weights appearing in Tables 9-11 are identical to those in Tables 6-8, respectively. 22 11 NERA Economic Consulting Tables 6-8, respectively).23 However, as we shall see below neither of these two conditions is true. For two reasons, the percentages in the three previous tables are not suitable as availability measures. First, it is likely that some proportion of the DBEs listed in the tables are not actually minority-owned or woman-owned. Second, it is likely that there are additional "unlisted" DBEs among all the businesses included in Tables 6-8. Such businesses do not appear in any of the directories we gathered and are therefore not included as DBEs in Tables 9-11. Additional steps are required to test these two conditions and to arrive at a more accurate representation of DBE availability in the State of Washington. We discuss these steps in Sections III.C and III.D below. C. Verify Listed DBEs and Estimate Unlisted DBEs It is likely that information on DBEs from MarketPlace and other DBE directories is not all correct. Phenomena such as ownership changes, associate or mentor status, recording errors, or even outright misrepresentation could lead to businesses being listed as DBEs in a particular directory even though they are actually owned by white males. Other things equal, this type of error would cause our availability estimate to be biased upward from the "true" availability number. The second likelihood that must be addressed is that not all DBE businesses are necessarily listed--either in MarketPlace or in any of the other directories we collected. Such phenomena as geographic relocation, ownership changes, directory compilation errors, and limitations in DBE outreach could all lead to DBEs being unlisted. Other things equal, this type of error would cause our availability estimate to be biased downward from the "true" availability number. In our experience, we have found that both types of bias are not uncommon. For this study, we attempted to correct for the effect of these biases using statistical sampling procedures. We surveyed a large stratified random sample of 1,501 relevant businesses by 23 The industry weights appearing in Tables 6-8 are identical to those in Tables 9-11. The "average availability" figure appearing at the bottom of each table is unweighted. That is, neither product market weights nor geographic weights have been applied. These weights are applied below. 12 NERA Economic Consulting telephone and measured how often they were misclassified (or unclassified) by race and/or sex.24 Strata were defined according to SIC code groups and listed DBE status.25 The survey was conducted by telephone during February and March 2005. Up to ten attempts were made to reach each business and speak with an appropriate respondent. Attempts were scheduled for a mix of day and evening, weekdays and weekends, and appointments were scheduled for callbacks when necessary. Of the 1,501 firms in our sample, 600 were listed DBEs and 901 were unclassified by race or sex. However, 331 establishments were excluded as "unable to contact." These resulted primarily from wrong phone numbers and phone numbers that had been disconnected or were no longer in service. Of the remaining 1,170 firms, 470 were listed DBEs and the remaining 700 establishments were unclassified. The first part of the survey tested whether our sample of listed DBEs was correctly classified by race and/or sex. The second part of the survey tested whether the unclassified firms could all be properly classified as non-DBEs. Both elements of the survey are described in more detail below. 1. Survey of Listed DBEs We selected a stratified random sample of 600 listed DBEs to verify the race and gender status of their owner(s). Of these, 130 (14.4%) were excluded as "unable to contact." Of the 470 remaining establishments, we obtained complete interviews from 353, for a response rate of 75.1 percent. 24 A very similar methodology has been employed by the Federal Reserve Board to deal with similar problems in designing and implementing the National Surveys of Small Business Finances for 1993 and 1998. See Catherine Haggerty, Karen Grigorian, Rachel Harter and John D. Wolken. "The 1998 Survey of Small Business Finances: Sampling and Level of Effort Associated with Gaining Cooperation from Minority-Owned Business," Proceedings of the Second International Conference on Establishment Surveys, Buffalo, N.Y., June 17-21, 2000. Five separate SIC strata were created according to industry weight. SIC codes with larger weights were sampled with higher probability. Together, these five strata account for more than 95 percent of all WSDOT contracting dollars. A sixth stratum was added to capture all remaining SIC codes. All six strata were then split according to listed DBE status to create a total of 12 strata. Generally, listed DBEs were sampled at a higher rate than unclassified establishments. 25 13 NERA Economic Consulting Of the 353 establishments interviewed, 75 (21.2%) were owned by White males. The amount of misclassification was substantial in every SIC stratum, and was highest in stratum 1 (SIC 16), as shown in Table 15. Misclassification was substantial as well in all Highway Regions, in the North Central Region in particular, as shown in Table 16. Misclassification varied by putative race and sex as well, and was highest among apparent White female firms, as shown in Table 17.26 The race and gender status of the listed DBEs responding to the survey was changed, if necessary, according to the survey results. For example, if a business originally listed as a White female DBE was actually owned by a White male, then that business was counted as a White male for purposes of calculating DBE availability. But what about the remaining putative White female-owned establishments that we did not interview? For these businesses, we must estimate their DBE status since we did not directly obtain it (because we did not interview them). We base our estimates on the amount of misclassification we observed among the White female-owned firms that we succeeded in interviewing. In this example, our interviews show that 62.2 percent of these firms are actually White female-owned, 30.6 percent are actually White male-owned, and 8.2 percent are actually minority-owned. Therefore, we assign each of the remaining putative White female firms a 62.2 percent probability of actually being White female-owned, a 30.6 percent probability of actually being White male-owned, and an 8.2 percent probability of being minority-owned. We repeated this procedure within each sample stratum and for all putative race and sex categories. 2. Survey of Unclassified Businesses In a manner exactly analogous to our survey of listed DBEs, in the second part of our survey we examined unclassified businesses, i.e. any business that was not originally identified as a DBE, either in MarketPlace or in one or more of the other directories collected for this study. 26 By "putative," we mean the race and sex that we initially assigned to each firm based on the information provided by Dun & Bradstreet or by our master M/W/DBE directory. 14 NERA Economic Consulting We selected a stratified random sample of 901 unclassified businesses to verify the race and gender status of their owner(s). Of these, 201 (22.3%) were excluded as "unable to contact." Of the 700 remaining establishments, we obtained 519 complete interviews, for a response rate of 74.1 percent. Of the 519 establishments interviewed, 460 (88.6%) were owned by White males, 33 (6.4%) by White females, and 26 (5.0%) by minorities. A similar phenomenon was observed within each stratum (Table 18) as well as within each highway district (Table 19). As with the survey of listed DBEs, the race and gender status of unclassified businesses was changed, if necessary, according to the survey results. For example, if an interviewed business that was originally unclassified indicated that they were actually owned by a White male, then that business was counted as a White male for purposes of the DBE availability calculation. If they indicated they were White female-owned, they were counted as White female, and so on. For unclassified businesses that were not interviewed, we assigned probability values (probability actually White male-owned, probability actually White female-owned, probability actually Black-owned, etc.) based on the interview responses. We again carried out the probability assignment procedure within each stratum. Clearly, the large majority of unclassified businesses (almost 89 percent overall) are White male-owned. Nevertheless, more than 11 percent were not White male-owned. Of the latter, the largest group was owned by White females, with descending size shares accounted for by Asians, Native Americans, Hispanics, and Blacks, respectively. Table 7C shows the actual survey results by race and sex. 15 NERA Economic Consulting IV. ESTIMATING BASELINE DBE AVAILABILITY All the steps necessary to calculate overall weighted average DBE availability are now complete. We briefly summarize each step below. Table 21 details the results from each step for all WSDOT federally-assisted contracting activity. Tables 22-23 repeat the process for construction and architectural/engineering design contracts. Identify the relevant geographic market. Determine the states and counties where prime contractors and subcontractors are located based on WSDOT's contract expenditure data. Identify the geographic areas that account for the majority of WSDOT's contract and subcontract activity. Identify the relevant product market and associated industry weights. Determine which SIC codes best represent contracting and subcontracting opportunities on WSDOT projects with federal participation, based on expenditure data for WSDOT's construction and architectural/engineering design contracts and subcontracts. Next, calculate the dollar value attributable to each SIC code as a percentage distribution. The resulting percentage figures are used to calculate industry-weighted DBE availability. In contrast to an unweighted figure, the industry-weighted DBE availability figure gives greater weight to DBE availability from those industries where WSDOT spends more contract dollars, and lesser weight to availability in those industries where fewer dollars are spent. Count all businesses in the relevant geographic and product market. Determine the total number of businesses in each relevant SIC code, state, and county from Dun & Bradstreet's MarketPlace. This determination was made overall as well as separately for construction and consulting. Identify "listed" DBE businesses in relevant markets. Some DBEs were directly identified in Dun & Bradstreet's MarketPlace or in WSDOT's DBE directory. Other businesses in MarketPlace were identified as DBEs by cross-referencing name and address information from numerous regional directories of minority- and women-owned firms collected for this study. This determination was made overall as well as separately for construction and consulting. 16 NERA Economic Consulting Verify ownership status of listed DBEs. To correct for race and sex misclassification, conduct interviews with listed DBEs to verify ownership status. Calculate the percentage of listed DBEs that are actually owned by White males. Separate calculations were made by SIC code grouping and by race and sex. Verify ownership status of unclassified firms. To correct for race and sex misclassification, conduct interviews with businesses that were not listed as DBEs in order to determine their ownership status. Calculate the percentage of unclassified businesses that are actually DBEs and non-DBEs. Separate calculations were made by SIC code grouping and by race and sex. Table 21 shows a total of 40,449 businesses operating in the 27 SIC codes within WSDOT's geographic market (see Table 6). Of these, 13.02 percent were listed DBEs. With industry weights, the percentage shrinks to 9.72 percent. This decrease occurs primarily because the proportion of listed DBEs in certain industries is less than the overall average. In particular, the proportion of listed DBEs in SIC 1611, at 8.95 percent, is substantially lower than the overall average of 13.02 percent. Our misclassification survey found that approximately 21 percent of listed DBEs were not actually DBEs (see tables 1517). Our survey also found that approximately 11 percent of unclassified firms were actually DBEs (see tables 18 & 19). Combining the results of these two surveys and applying them as probability weights to the baseline business population yields an unweighted DBE availability of 28.21 percent, which then falls significantly to the final overall baseline availability figure of 18.77 percent once industry weights are applied. Tables 22-23 provide similar derivations for construction and consulting, respectively. The final results of our baseline DBE availability analysis for WSDOT are shown in Table 24. Overall, DBE availability for WSDOT contracts is estimated to be 18.77 percent. Availability for construction contracts is estimated to be 19.59 percent. Availability for consulting contracts is estimated to be 14.88 percent. Availability results are also presented by highway regions and by the race and sex of business ownership. 17 NERA Economic Consulting V. DBE PARTICIPATION IN WSDOT CONTRACTING AND SUBCONTRACTING, FFY 1999-2003 Using the databases of WSDOT contracting and subcontracting activity described above in Section II.A.1 and II.A.2, we calculated the fraction of all contracts, subcontracts, contract dollars and subcontract dollars received by DBEs. Tables 25-36 below provide this information from several important perspectives: (1) federallyfunded versus non-federally-funded, (2) prime contract gross amount versus prime contract amount net of subcontracted amounts, (3) prime contract dollars versus prime contract and first-tier subcontract dollars combined. Tables 25-30 cover WSDOT construction projects and Tables 31-36 cover WSDOT consulting projects. Results are presented for White males, White females, Blacks, Asians, Native Americans, all MBEs combined, and all DBEs combined. An examination of the results in Tables 25-36 shows that: (1) the DBE share of contracts is greater than the DBE share of contract dollars, (2) DBE participation in subcontracting is greater than DBE participation in prime contracting, and (3) in Construction, DBE participation is much higher on federally-funded projects than on non-federally-funded projects. The amount of DBE participation that could be expected in the absence of raceor sex-conscious goals can be estimated based on the amount of DBE participation of projects without DBE goals. As a proxy for this, we consider DBE participation on non-federally-funded contracts and subcontracts, as shown in Tables 30 and 36. Table 30 shows that DBE participation on non-federally-funded construction contracts and subcontracts during the review period was 2.97 percent. For consulting, the figure is 10.66 percent (Table 36). 18 NERA Economic Consulting VI. DISPARITIES IN MBE BUSINESS FORMATION AND BUSINESS OWNER EARNINGS In this Study, we examine qualitative and quantitative evidence relevant to establishing whether expected DBE availability in WSDOT's construction and consulting contracting markets would, absent business-related discrimination, be substantially and significantly higher or lower than the levels shown above in Table 24. The baseline availability figures calculated in the previous section represent the percentage of businesses in WSDOT's construction and consulting markets that are owned by minorities and/or women. These availability figures will be artificially low if discrimination has led minorities and women to be more reluctant to start businesses or if it has contributed to the businesses they start being less profitable and therefore more likely to fail. For this reason, 49 CFR ?26.45 requires recipients of federal funds to consider whether an adjustment to the baseline DBE availability figures such as those reported in Table 9 would be necessary in order to approximate the amount of DBE availability that would be expected in a race-neutral marketplace, that is, "but for" discrimination. This is referred to in the regulations as the step 2 adjustment. 27 Specifically, recipients must examine the volume of work DBEs have performed for them in the past as well as findings from any relevant disparity studies conducted within the recipient's jurisdiction. Recipients must also consider "evidence from related fields that affect the opportunities for DBEs to form, grow and compete" to the extent available.28 In keeping with these requirements, this final section of the Study summarizes evidence relevant to whether an adjustment is warranted and, if so, what size adjustment would be narrowly tailored to that evidence. First, we review the microeconomic and microeconometric literature on self-employment and entrepreneurship. Secondly, we present statistical evidence of disparities in business formation and business owner earnings, based on entrepreneur microdata from the 2000 Decennial Census and from the 1979-2002 Current Population Surveys. The presence of 27 28 49 CFR ? 26.45(d). 49 CFR ? 26.45(d)(2). 19 NERA Economic Consulting statistically significant business formation and earnings disparities is consistent with present discrimination in the WSDOT marketplace and/or the present effects of past discrimination in the WSDOT marketplace. This evidence of business formation disparities forms the basis for quantifying the amount of upward or downward adjustment from Step 1 availability that would be consistent with a race-neutral marketplace. Finally, in order to shed light on how much of WSDOT's annual DBE goal is susceptible to fulfillment by race-neutral measures alone, we examine the past volume of construction and consulting work performed for WSDOT and its prime contractors by DBEs, comparing utilization differences on federally-funded versus non-federally funded projects as well as differences on projects with DBE goals versus projects without DBE goals. NERA's estimates of DBE availability from the previous section (See Table 24) are substantially higher than average DBE utilization levels achieved by WSDOT between FFY 1999 and FFY 2003. 29 A. Review of Relevant Literature We examine here disparities in business formation and earnings principally in the private sector, where contracting and procurement activity is generally not subject to MWBE requirements. Statistical examination of disparities in the private sector economy surrounding the State of Washington is important for at least three reasons. First, to the extent that discriminatory practices by contractors, suppliers, insurers, lenders, customers, and others limit the ability of DBEs to compete, those practices are likely to be felt in the larger private sector as well as in the public sector. Second, examining the utilization of DBEs in the private sector provides an indicator of the extent to which DBEs are used in the absence of affirmative action efforts, since few firms in the private sector make such efforts. Third, the Supreme Court in Croson acknowledged that state and local governments have a constitutional duty not to contribute to the perpetuation of racial or ethnic discrimination in the private sector of the local economy. 29 See Tables 25-36. 20 NERA Economic Consulting After years of comparative neglect, research on the economics of entrepreneurship--especially upon self-employment--is beginning to expand.30 In the U.S. for example, minorities start businesses at much lower rates than non-Hispanic whites. These disparities persist even when factors such as geography, industry, occupation, age, education and assets are held constant. 31 One possible impediment to entrepreneurship among minorities is lack of capital. 32 The key test shows that, all else remaining equal, people with greater family assets are more likely to switch to self-employment from employment. This asset variable enters probit equations significantly and with a quadratic form. Indeed, the probability of selfemployment depends positively upon whether the individual ever received an inheritance or gift.33 Further, house prices through the impact on equity withdrawal play a powerful role in affecting the supply of small new firms.34 Again this is suggestive of capital constraints.. Transfers of firms within families will also help to preserve the status quo and work against 30 Blanchflower [8]. Microeconometric work includes Fuchs [30], Borjas and Bronars [17], Evans and Jovanovic [22], Evans and Leighton [23], Fairlie [24], Fairlie and Meyer [11, 26], Reardon [48], Wainwright for the United States [54], Rees and Shah [49], Pickles and O'Farrell [46], Blanchflower and Oswald [11, 12, 13], Meager [43], Taylor [53], Robson for the UK[50, 51] , DeWit and van Winden for the Netherlands [21], Alba-Ramirez for Spain [2], Bernhardt [6], Schuetze [52], Arai [3], Lentz and Laband [40], and Kuhn and Schuetze] for Canada [38, Laferrere and McEntee for France [39], Blanchflower and Meyer [10] and Kidd for Australia [36], and Foti and Vivarelli for Italy [29]. There are also several theoretical papers including Kihlstrom and Laffonte [36], Kanbur [35], Coate and Tennyson [19], and Holmes and Schmitz [31], plus a few papers that draw comparisons across countries i.e. Schuetze for Canada and the U.S. [52], Blanchflower and Meyer for Australia and the U.S. [10], Alba-Ramirez for Spain and the United States [2], and Acs and Evans for many countries [1]. Public Use Microdata Samples (PUMS) data from the 1990 Census, Wainwright [54]. In work based on U.S. micro data at the level of the individual, Evans and Leighton [23], and Evans and Jovanovic [22] have argued formally that entrepreneurs face liquidity constraints. The authors use the National Longitudinal Survey of Young Men for 1966-1981, and the Current Population Surveys for 1968-1987. Blanchflower and Oswald [12]. This emerges from British data, the National Child Development Study; a birth cohort of children born in March 1958 who have been followed for the whole of their lives. They also find that, when directly questioned in interview surveys, potential entrepreneurs say that raising capital is their principal problem. Additionally, Blanchflower and Oswald find that the selfemployed report higher levels of job and life satisfaction than employees, and that psychological test scores play only a small role in explaining entry into self-employment. Work by Holtz-Eakin, Joulfaian and Rosen drew similar conclusions using different methods on U.S. data [32, 33]. Black, Meza, and Jeffreys [7]; Cowling and Mitchell [20]. 31 32 33 34 21 NERA Economic Consulting the interests of Blacks in particular who do not have as strong a history of business ownership as indigenous whites. Analogously, because the offspring of self-employed fathers are more likely than others to become self-employed the historically low rates of self-employment among Blacks and Latinos may contribute to their low contemporary rates.35 Nationally, the self-employment rate of Black males is one third of that of White males and has remained roughly constant since 1910. Neither trends in demographic factors, including the Great Migration and the racial convergence in education levels, nor an initial lack of business experience, nor the lack of traditions in business enterprise among blacks that resulted from slavery can explain a substantial part of the current racial gap in self-employment".36 A considerable part of the explanation of the differences between the Black and White self-employment rate can be attributed to discrimination.37 There is strong evidence that racial differences in levels of financial capital have significant effects upon racial patterns in business failure rates. 38 Further, the black exit rate from self-employment is twice as high as that of whites.39 B. Race and Sex Disparities in Earnings In this section we examine earnings to determine whether minority and female entrepreneurs earn less from their businesses than do their White male counterparts. Other things equal, if minority and female business owners as a group cannot achieve comparable earnings from their businesses as similarly-situated non-minorities because of discrimination, then failure rates for MWBEs will be higher and MWBE formation rates will be lower than would be observed in a race- and sex-neutral marketplace. Both phenomena would contribute directly to lower levels of minority and female business ownership. 35 36 37 38 39 Hout and Rosen [34]. Fairlie and Meyer (2000) ([27] p. 664) Robert Fairlie [24] and Wainwright [54]. Tim Bates [5]. Fairlie [24]. 22 NERA Economic Consulting Below, we first examine earnings disparities among wage and salary employees, that is, non-business owners. It is critical to examine this segment of the labor force since a key source of new entrepreneurs in any given industry is the pool of experienced wage and salary workers in that same industry.40 Any employment discrimination that adversely impacts the ability of minorities or women to succeed in the labor force directly shrinks the available pool of potential MWBEs. In almost every instance examined, a statistically significant adverse impact on earnings is observed in both the economy at large and in the construction and construction-related professional services sector.41 We then turn to an examination of differences in earnings among the self-employed, that is, among business owners. Here too, among the pool of minorities and women who have formed businesses despite discrimination in both employment opportunities and business opportunities, statistically significant adverse impacts are observed in the vast majority of cases both in construction and the economy as a whole. The remainder of this section discusses the methods and data we employed and presents the specific findings we obtained. 1. Methods We used a statistical technique known as linear regression analysis to estimate the effect of each of a set of observable characteristics, such as education and age, on an outcome variable of interest. In this case, the outcome variable of interest is earnings and we used regression to compare earnings among individuals in similar geographic and product markets at similar points in time and with similar years of education and potential labor market experience and see if any adverse race or sex differences remain. In a 40 41 Blanchflower [8, 9]. There is a growing body of evidence that discriminatory constraints in the capital market prevent minority-owned businesses from obtaining business loans. Furthermore, even when they are able to obtain them there is evidence that these loans are not obtained on equal terms: minority-owned firms have to pay higher interest rates, other things being equal. We have written in other studies regarding racial discrimination in commercial credit and capital markets throughout the U.S. This is another form of discrimination with an obvious and direct impact on the ability of racial minorities to form businesses and to expand or grow previously formed businesses. Additionally, see the detailed discussion of this phenomenon in D. G. Blanchflower, P. B. Levine, and D. Zimmerman, "Discrimination in the market for small business credit market", NBER Working Paper W6840, 1999. 23 NERA Economic Consulting discrimination free market place, one would not expect to observe significant differences in earnings by race or sex among such similarly situated observations. Regression also allows us to narrowly tailor our statistical tests to the State of Washington and assess whether disparities in the State of Washington are statistically significantly different from those observed elsewhere in the nation. Starting from an economy-wide data set, we first estimate the basic model of earnings differences just described and also include an indicator variable for the State of Washington. This model appears as Specification (1) in Tables 37 through 48. Next, we estimate Specification (2), which is the same model as (1) but with the addition of indicator variables that interact race, sex, and the State of Washington. Specification (3) represents our ultimate specification, which includes all the variables from the basic model as well as any of the interaction terms from Specification (2) that were statistically significant.42 Any negative and statistically significant differences by race or sex that remain in Specification (3) after holding all of these other factors constant--time, age, education, geography, and industry--are consistent with what would be observed in a market suffering from business-related discrimination. 2. Data The analyses undertaken in this report require individual-level data (i.e. "microdata") with relevant information on business ownership status and other key socioeconomic characteristics. Two primary sources of such data are available. The first is the Five Percent Public Use Microdata Samples (PUMS) from the 2000 decennial census. The 2000 PUMS contains observations representing five percent of all U.S. housing units and the persons in them (approximately 14 million records). Released in late 2003, the PUMS provides the full range of population and housing information collected in the 2000 census. Business ownership status is identified in the PUMS through the "class of worker" variable, which distinguishes the unincorporated and incorporated self-employed from others in the labor force. The presence of the class of worker variable 42 If none of these terms is significant then Specification (3) reduces to Specification (1). 24 NERA Economic Consulting allows us to construct a detailed cross-sectional sample of individual business owners and their associated earnings. The second source of data is the Current Population Survey (CPS). The CPS has been conducted monthly by the Census Bureau and the Bureau of Labor Statistics for over 40 years, and is a primary source of official government statistics on employment and unemployment. Currently, about 56,500 households are scientifically selected for the CPS on the basis of area of residence in order to represent the nation as a whole, individual states and the largest metropolitan areas. In addition to information on employment status, the CPS collects information on age, sex, race, marital status, educational attainment, earnings, occupation, industry, and other characteristics. These statistics serve to update the information collected every 10 years through the decennial census.43 3. Findings: Race and Sex Disparities in Wage and Salary Earnings Tables 37 through 42 report results from our regression analyses of annual earnings among wage and salary workers. Tables 37 through 39 focus on the economy as a whole and Tables 40 through 42 on construction and construction-related professional services. Tables 37 and 40 are derived from the 2000 PUMS, Tables 38 and 41 are derived from the 1979-1991 CPS, and Tables 39 and 42 are derived from the 1992-2002 CPS. The numbers shown in each of these six tables indicate the percentage difference between the average wages of a given race/sex group and comparable White males. 43 Since 1979, about a quarter of the households in each monthly CPS survey have been asked to provide additional information, including usual weekly earnings and weekly hours of work. These households are said to be in "Outgoing Rotation Groups" (ORG) because of the way the CPS rotates households for interviews. Each household selected for the survey is interviewed once a month for four consecutive months, not interviewed for eight months, and interviewed again once a month for four more months. The households in the ORG are those that are in either the fourth or the eighth survey. The ORG files of the CPS include individual data for about 30,000 individuals each month, or over 350,000 per year. Data in which the State of Washington is identifiable are available in a comparable form from 1986 through 2002. Data from the ORG files are used below in addition to the PUMS to examine earnings disparities among wage and salary workers. The ORG files however, do not contain data on the earnings of the self-employed. Annual earnings, whether from wages or self-employment are available from the March CPS, however, also known as the Annual Demographic File. This latter file also contains the basic monthly demographic and labor force data. In the March CPS, data on employment, earnings, and income refer to the preceding year, although demographic data refer to the time of the survey. The March surveys are therefore included for the years 1987-2003. Because the information (continued...) 25 NERA Economic Consulting a. Specification (1) - the Basic Model For example, in Table 37 Specification (1) the estimated percentage difference in annual wages between Blacks (both sexes) and White males in 2000 was -30.4 percent. That is, average annual wages among Blacks were 30.4 percent lower than for White males who were otherwise similar in terms of geographic location, industry, age, and education. The number in parentheses below each percentage difference is the t-statistic, which indicates whether the estimated percentage difference is statistically significant or not. In Tables 37 through 42, a t-statistic of 1.99 or larger indicates statistical significance at a 95 percent confidence level or better.44 In the example just used, the t-statistic of 197.61 indicates that the result is statistically significant. Specification (1) in Tables 37-39 shows negative and statistically significant wage disparities for Blacks, Hispanics, Asians, Native Americans, persons reporting in multiple race categories, and White women consistent with the presence of discrimination in these markets. Observed disparities are large as well, ranging from a low of -16.7 percent for Hispanics in Table 38 to a high of -35.7 percent for White women in Table 37. Specification (1) in Tables 40 through 42 shows similar results when the basic analysis is restricted to the construction and construction-related professional services sector. In this sector, large, negative, and statistically significant wage disparities are observed for all minority groups and for white women. For Blacks, the large wage disparities observed in the construction sector are similar to those observed economy-wide. Large wage disparities in construction are also observed for Hispanics, Asians, and Native Americans; however, the differences are smaller than those observed in the economy as a whole. For White women, large disparities are observed both economy-wide and in construction--however, disparities in construction are larger. (...continued) relates to the preceding year, the earnings data relate to the years 1986-2002. The sample consists of any individual who reports positive self-employment earnings in the year preceding the interview. 44 From a two-tailed test. 26 NERA Economic Consulting Specification (1) in, respectively, Tables 38 and 39 and Tables 41 and 42 describes changes in observed wage disparities over time. For the economy as a whole, as well as for the construction sector, disparities for Blacks became slightly smaller between 1979-1991 (Tables 38 and 41) and 1992-2002 (Table 39 and 42), but remain large (average wages more than 20 percent below comparable White males). For Hispanics, wage disparities increased substantially during the same period and average wages remain 14-20 percent lower than for comparable White males in construction and elsewhere. For White women, wage disparities grew substantially smaller between the two periods, both in construction and in the economy as a whole, although they remain large (average wages 18-25 percent below comparable White males).45 Finally, the indicator variable for the State of Washington is positive and statistically significant in the 2000 PUMS data, although this is not the case in the CPS data. The PUMS data indicate that residents of the State of Washington enjoy, on average, a modest wage advantage over their similarly situated counterparts elsewhere in the nation. Unfortunately, the observed wage advantages fail to offset the much larger wage disadvantages observed for minorities and women throughout the nation and the State of Washington. b. Specifications (2) and (3) - the Full Model Including WashingtonSpecific Interaction Terms Next, we turn to Specifications (2) and (3) in Tables 37 through 42. In each of these Tables, Specification (2) is the basic regression model enhanced by the addition of a set of interaction terms that test whether minorities and women in the State of Washington differ significantly from those elsewhere in the U.S. economy. Specification (2) in Table 37, for example, shows once again the -30.5 percent wage difference that estimates the direct effect of being Black in 2000, as well as a statistically significant 10.3 percent wage increment in that year that captures the indirect effect of residing in the State of Washington and being Black. Therefore, the net wage disparity for Blacks in the State of Washington is approximately -20.2 percent (-30.5 percent plus 10.3 percent). 45 It is not possible to perform a similar comparison for Asians or Native Americans, as they were not (continued...) 27 NERA Economic Consulting Specification (3) simply repeats Specification (2), dropping any Washington interaction terms that are not statistically significant. In Table 39, for example, the only interaction terms included in the final specification were for Blacks and Asians. The net result of Specification (3) in Tables 37, 38 and 39 is evidence of large, negative and statistically significant wage disparities for all minority groups and for White women. The same result holds in Construction and Consulting ( Tables 40, 41, and 42). Clearly, prime age minorities and women earn substantially and significantly less from their labors than their White male counterparts. Such disparities are symptoms of discrimination in the labor force that, in addition to its direct effect on workers, reduce the future availability of DBEs by stifling opportunities for minorities and women to progress through precisely those internal labor markets and occupational hierarchies that are most likely to lead to entrepreneurial opportunities in the first place. These disparities reflect more than mere "societal discrimination" because they demonstrate the relationship between discrimination in the job market and reduced entrepreneurial opportunities for minorities and women. Other things equal, these reduced entrepreneurial opportunities in turn lead to lower DBE availability levels than would be observed in a race- and sex-neutral marketplace. 4. Findings: Race and Sex Disparities in Business Owner Earnings We turn next to the analysis of race and sex disparities in business owner earnings. Tables 43 through 48 report results from regression analyses of earnings from selfemployment. Tables 43 through 45 focus on the economy as a whole and Tables 46 through 48 on construction and construction-related professional services. Tables 43 and 46 are derived from the 2000 PUMS, Tables 44 and 47 are derived from the 1979-1991 CPS, and Tables 45 and 48 are derived from the 1992-2002 CPS. The numbers shown in each of these six tables indicate the percentage difference between the average annual selfemployment earnings of a given race/sex group and comparable White males. (...continued) identified separately in the CPS prior to 1992 and instead were classified together as "Other Race." 28 NERA Economic Consulting a. Specification (1) - the Basic Model Specification (1) in Tables 43 through 45 shows negative and statistically significant and large wage disparities for Blacks, Hispanics, Asians, Native Americans, persons of mixed race, and White women consistent with the presence of discrimination in these markets. The measured difference for Blacks ranges between 30 percent and 59 percent; for Hispanics, from 19 percent to 39 percent; for Asians, from 4 percent to 22 percent; and for Native Americans, from 38 percent to 51 percent. The largest business owner earnings disparities, however, are observed for White women: between 44 percent and almost 73 percent. Specification (1) in Tables 46 through 48 shows similar results for the construction and construction-related professional services sector. Large negative earnings disparities are observed in every case--in particular for Blacks and White Females. Most of instances these differences are also statistically significant. Specification (1) in, respectively, Tables 44 and 45 and Tables 47 and 48 describes changes in observed business owner earnings disparities over time. For the economy as a whole as well as for the construction sector, large disparities for Blacks increased between 1979-1991 (Tables 44 and 47) and 1992-2002 (Table 45 and 48). For Blacks and Hispanics, in the economy as a whole, the large earnings disparities observed in the 1979- 1991 period grew even larger from 1992-2002. In the construction sector, disparities for both groups remained large but were smaller in 1992-2002 than in 1979-1991. For White women, while disparities have lessened somewhat in the economy as a whole, in the construction sector disparities remain among the largest observed (between 50 percent and 85 percent lower than White males). Finally, with respect to Specification (1), the indicator variable for the State of Washington is insignificantly different from zero 4 of 6 times in Tables 43-48. In the two cases in which it is statistically significant, it is negative. This indicates that residents of the State of Washington enjoy no apparent earnings advantage over similarly situated entrepreneurs elsewhere in the nation, and might in fact be at somewhat of an earnings disadvantage. 29 NERA Economic Consulting b. Specifications (2) and (3) - the Full Model Including WashingtonSpecific Interaction Terms Next we turn to Specifications (2) and (3) in Tables 43 through 48. Specification (2) is the basic regression model enhanced by a set of interaction terms to test whether minorities and women in the State of Washington differ significantly from persons elsewhere in the U.S. economy. Specification (3) drops any Washington interaction terms that are not statistically significant. For the economy as a whole (Tables 43 through 45), none of the Washington interaction terms is statistically significant, indicating that estimates for Washington are similar to results from elsewhere in the nation. The final results for these three tables therefore are complied in Specification (1). The same is true in Tables 46 and 48, though not Table 47, where the final results are as in Specification (3). As was the case for wage and salary earners, prime age minority and female entrepreneurs earn substantially and significantly less from their efforts than similarly situated White male entrepreneurs. These disparities are a symptom of discrimination in commercial markets that directly and adversely affects DBEs. Other things equal, if minorities and women cannot earn remuneration from their entrepreneurial efforts comparable to that of White males, growth rates will slow, business failure rates will increase, and as demonstrated in the next section, business formation rates will decrease. Combined, these phenomena result in lower DBE availability levels than would be observed in a race- and sex-neutral marketplace. C. Race and Sex Disparities in Business Formation Finally, we turn to the analysis of race and sex disparities in business formation.46 In this section, we compare self-employment rates by race and sex to determine whether minorities or women are as likely to enter the ranks of entrepreneurs as similarly-situated White males. We find that they are not as likely to do so and that minority business 46 We use the phrases "business formation rates" and "self-employment rates" interchangeably in this Study. 30 NERA Economic Consulting formation rates would likely be substantially and significantly higher if markets operated in a race- and sex-neutral manner. Discrimination in the labor market, symptoms of which are evidenced in Section B.3 above, might cause wage and salary workers to turn to self-employment in hopes of encountering less discrimination from customers and suppliers than from employers and coworkers. Other things equal, and assuming minority and female workers did not believe that discrimination pervaded commercial markets as well, this would lead minority and female business formation rates to be higher than would otherwise be expected. On the other hand, discrimination in the labor market prevents minorities and women from acquiring the very skills, experience, and positions that are often observed among those who leave the ranks of the wage and salary earners to start their own businesses. Many construction contracting concerns have been formed by men who were once employed as foreman for other contractors, fewer by those who were employed instead as laborers. Similarly, discrimination in commercial capital and credit markets, as well as asset and wealth distribution, prevents minorities and women from acquiring the financial credit and capital that are so often prerequisite to starting or expanding a business. Other things equal, these phenomena would lead minority and female business formation rates to be lower than otherwise would be expected. Further, discrimination by commercial customers and suppliers against DBEs, symptoms of which are evidenced in Section B.4 above and elsewhere, operates to increase input prices and lower output prices for DBEs. This discrimination leads to higher rates of failure for some minority and women firms, lower rates of profitability and growth for others, and prevents some minorities and women from ever starting businesses.47 All of these phenomena, other things equal, would contribute directly to lower observed rates of minority and female self-employment. 47 See also the materials cited at fn. 41 supra. 31 NERA Economic Consulting 1. Methods and Data To see if minorities or White women are as likely to be business owners as are comparable White males, we use a statistical technique known as Probit regression. Probit regression is used to determine the relationship between a categorical variable--one that can be characterized in terms of a yes or no response as opposed to a continuous number-- and a set of characteristics that are related to the outcome of the categorical variable. Probit regression produces estimates of the extent to which each characteristic is positively or negatively related to the likelihood that the categorical variable will be a yes or no. For example, Probit regression is used by statisticians to estimate the likelihood that an individual participates in the labor force, retires this year, or contracts a particular disease-- these are all variables that can be categorized by a response of yes (for example, she is in the labor force) or no (for example, she is not in the labor force)--and the extent to which certain factors are positively or negatively related to the likelihood (for example, the more education she has, the more likely that she is in the labor force). Probit regression is one of several techniques that can be used to examine qualitative outcomes. Generally, other techniques such as Logit regression yield similar results.48 In the present case, Probit regression is used to examine the relationship between the choice to own a business (yes or no) the other demographic and socioeconomic characteristics in our basic model. The underlying data for this section is once again the 2000 PUMS, the 1979-1991 CPS, and the 1992-2002 CPS. 2. Findings: Race and Sex Disparities in Business Formation As a point of reference for what follows, Tables 49 and 50 provide a summary of business ownership rates in 2000 by race and sex. A striking feature of both tables is how much higher business ownership rates in the United States are for White males than for any other group. Table 49, for example, shows almost an 8 percentage point difference between the overall self-employment rate of Blacks and White Males in the State of Washington 48 For a detailed discussion, see G.S. Maddala, Limited Dependent and Qualitative Variables in Econometrics, Cambridge University Press, 1983. Probit analysis is performed here using the "dprobit" command in the statistical program STATA. 32 NERA Economic Consulting (13.3 - 5.7 = 7.6), and Table 50 shows more than a 16 point difference in the construction sector self-employment rate for this group. Results such as this are observed whether we consider the country as a whole or only the State of Washington, it is apparent in the construction sector as well as in the economy as a whole, and it is evident for all minority groups and for White women. There is no doubt that part of the group differences shown in Tables 49 and 50 are associated with differences in the distribution of individual characteristics and preferences between minorities, women, and White males. It is well known that personal earnings tend to increase with age, for example. It is also true that the propensity toward self-employment increases with age.49 Since most minority populations in the U.S. have a lower median age than the non-Hispanic white population, we must examine whether the disparities in business ownership evidenced in Tables 49 and 50 are largely--or even entirely--due to differences in the age distribution of minorities compared to non-minorities or other factors such as education, geographic location, or industry preferences. The remainder of this section presents a series of regression analyses designed to address whether large, negative and statistically significant race and sex disparities are found among otherwise similarly-situated individuals. Tables 51 through 56 report results from regression analyses of the decision to start a business. Tables 51 through 53 focus on the economy as a whole and Tables 54 through 56 focus on construction and constructionrelated professional services. As in previous sections, the first in each triad of Tables is derived from the 2000 PUMS, the second from the 1979-1991 CPS, and the third from the 1992-2002 CPS. The numbers shown in each of these tables indicate the percentage point difference between the probability of self-employment for a given race/sex group and for comparable White males. a. Specification (1) - the Basic Model Specification (1) in Tables 51 through 53 shows negative, statistically significant and large business formation disparities for Blacks, Hispanics, Asians, Native Americans, 49 Wainwright [54] p. 86. 33 NERA Economic Consulting persons of mixed race, and White women consistent with the presence of discrimination in these markets. Specification (1) in Tables 54 through 56 shows similar large, negative, and statistically significant business formation disparities for every group in the construction and construction-related professional services sector. Once again, Specification (1) in, respectively, Tables 52 and 53 and Tables 55 and 56 describes changes in observed business owner earnings disparities over time. For the economy as a whole as well as for the construction sector, disparities for Blacks and Hispanics have actually worsened in recent years, while those for Asians and Native Americans have changed only little. In the construction sector, disparities for White women have lessened substantially in the construction sector, although they remain large. Disparities for White women in the economy as a whole, in contrast, did not change much between the two periods. Lastly, with respect to Specification (1), results on the indicator variable for the State of Washington indicate a positive self-employment effect relative to the rest of the nation in the 2000 PUMS data. b. Specifications (2) and (3) - the Full Model Including WashingtonSpecific Interaction Terms Several of the Washington interaction terms included in Specification (2) were significant. The final results are in Specification (3) for Tables 51-54, and in Specification (1) for Tables 55-56. To summarize for the economy-wide results (Tables 51-53): o The remaining difference for Blacks ranges between 1.6 and 4.7 percentage points (approximately 30-35 percent lower than the corresponding White male business formation rate).50 50 Because the overall White male self-employment rate is 13.6 percent (Table 49), the rate for comparable Blacks is approximately 30-35 percent lower than expected (i.e. 3.7 ? 13.6 0.27; 4.8 ? 13.6 0.35). 34 NERA Economic Consulting o For Hispanics, from 2.8 to 6.3 percentage points (approximately 21-46 percent lower than the White male business formation rate). o For Asians, from -0.5 to +0.4 percentage points (approximately 4 percent lower to 3 percent higher than the White male business formation rate). o For Native Americans, from 3.0 to 3.3 percentage points (approximately 22-24 percent lower than the White male business formation rate). o For White women, from 0.2 to 1.3 percentage points (approximately 1-10 percent lower than the White male business formation rate). To summarize for the construction sector results (Tables 54-56): o For Blacks, the remaining difference ranges between 8.5 to 19.9 percentage points (approximately 34-80 percent lower than the corresponding White male business formation rate). o For Hispanics, from 6.5 to 9.1 percentage points (approximately 26-36 percent lower than the White male business formation rate). o For Asians, from 5.6 to 7.5 percentage points (approximately 22-30 percent lower than the White male business formation rate). o For Native Americans, from 7.6 to 8.9 percentage points (approximately 30-36 percent lower than the White male business formation rate). o For White women, from 4.8 to 9.9 percentage points (approximately 19-40 percent lower than the White male business formation rate). c. Conclusions This section has demonstrated that observed DBE availability levels in the State of Washington are substantially and statistically significantly lower than those that would be observed if commercial markets operated in a race- and sex-neutral manner. This suggests that minorities and women are substantially and significantly less likely to own their own businesses as the result of discrimination than would be expected based upon their observable characteristics including age, education, geographic location, and industry. 35 NERA Economic Consulting These groups also suffer substantial and significant earnings disadvantages relative to comparable White males whether they work as employees or as entrepreneurs. 36 NERA Economic Consulting D. Estimates of Adjusted DBE Availability The Probit regression results for the Washington construction and architecture/engineering sector from Table 54 are combined with weighted average selfemployment rates by race and sex from the 2000 PUMS (Table 50) to determine the expected difference between baseline availability and expected availability in a race-neutral marketplace. These figures appear in column (2) of Table 57. Overall, the self-employment rate for minorities and women is 14.4 percent. According to the regression specification underlying Table 57, that rate would be 20.7 percent, or 43.8 percent higher, in a race and sex neutral marketplace. Put differently, the disparity ratio of the actual compared to the potential business formation rate is 0.70. Disparity ratios are large and statistically significant for all groups examined. The largest disparity observed is for Blacks (0.22), followed in descending order by that for Hispanics (0.58), Native Americans (0.64), Asians (0.71), Multiple races (0.77), and White females (0.78). Given the large disparities observed throughout Table 57, adjusted baseline estimates of DBE availability may be warranted to account for the continuing effects of discrimination, as directed by 49 CFR ? 26.45(d)(1)(ii). It is important to note, however, that even the unadjusted baseline DBE availability figure is substantially higher than the average level of DBE utilization that WSDOT has achieved in recent fiscal years.51 Finally, Table 58 summarizes our estimates of baseline DBE availability and adjusted DBE availability for construction and consulting, separately as well as combined. 51 See Section V, above. 37 FI A VII. TABLES 38 NERA Economic Consulting Table 1. Product Market for All WSDOT Contracts SIC Code SIC Description Percentage Cumulative Percentage 1611 1622 8711 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 0782 1799 7389 1795 7353 4212 1711 1781 4959 8743 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Engineering Services Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Public Relations Services Surveying Services TOTAL ($) 39.207 15.220 6.528 5.383 5.342 4.829 3.687 2.642 2.179 1.888 1.738 1.597 1.586 1.299 1.141 1.130 0.904 0.834 0.606 0.448 0.356 0.342 0.275 0.248 0.226 0.191 0.177 1,605,950,845 39.207 54.426 60.954 66.337 71.679 76.508 80.195 82.837 85.016 86.904 88.642 90.240 91.826 93.125 94.265 95.395 96.299 97.133 97.738 98.187 98.542 98.884 99.159 99.406 99.633 99.823 100.000 39 NERA Economic Consulting Table 2. Product Market for WSDOT Construction Contracts SIC Code SIC Description Percentage Cumulative Percentage 1611 1622 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 0782 1799 7389 1795 8711 7353 4212 1711 1781 4959 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Engineering Services Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Surveying Services TOTAL ($) 41.867 16.252 5.748 5.705 5.157 3.937 2.821 2.327 2.016 1.856 1.706 1.694 1.388 1.218 1.207 0.965 0.890 0.647 0.479 0.419 0.380 0.365 0.292 0.264 0.241 0.159 1,503,894,094 41.867 58.120 63.867 69.572 74.729 78.667 81.488 83.814 85.831 87.687 89.393 91.086 92.474 93.692 94.898 95.863 96.753 97.400 97.879 98.298 98.677 99.042 99.335 99.599 99.841 100.000 40 NERA Economic Consulting Table 3. Product Market for WSDOT Consulting Contracts SIC Code SIC Description Percentage Cumulative Percentage 8711 8743 8741 8748 8713 8742 Engineering Services Public Relations Services Management Services Business Consulting, n.e.c. Surveying Services Management Consulting Services TOTAL ($) 93.163 2.896 2.308 0.863 0.415 0.353 105,766,945 93.163 96.060 98.368 99.231 99.647 100.000 41 NERA Economic Consulting Table 4. Distribution of WSDOT Contract Dollars by Category Location Inside Washington Outside Washington Metropolitan Non-Metropolitan Northwest Region Olympic Region Eastern Region South Central Region North Central Region Southwest Region Outside WA Seattle-Bellevue-Everett, WA Bremerton, WA Spokane, WA Tacoma, WA Yakima, WA Portland-Vancouver, OR-WA Richland-Kennewick-Pasco, WA Bellingham, WA Eugene-Springfield, OR Olympia, WA Salem, OR San Diego, CA San Jose, CA San Francisco, CA All other metropolitan areas combined Non-metropolitan areas Construction (%) 93.7% 6.3% 89.1% 10.9% 44.4% 24.6% 11.9% 7.0% 3.1% 2.8% 6.2% 41.1% 14.4% 11.3% 6.5% 3.9% 3.5% 2.0% 1.9% 1.4% 1.2% 0.6% 0.0% 0.3% 0.0% <1.0% 10.9% Consulting (%) 92.4% 7.6% 99.0% 1.0% 87.6% 3.1% 0.6% <0.1% 1.0% 0.1% 7.6% 87.3% <0.1% 0.6% 1.6% <0.1% 0.2% 0.0% 0.3% 0.1% 1.4% 0.0% 2.9% 0.0% 1.4% 3.1% 1.0% Combined (%) 93.6% 6.4% 89.7% 10.3% 47.2% 23.2% 11.1% 6.5% 3.0% 2.7% 6.3% 44.2% 13.5% 10.6% 6.1% 3.7% 3.3% 1.8% 1.8% 1.3% 1.2% 0.5% 0.3% 0.3% 0.1% 1.0% 10.3% Source: NERA calculations from WSDOT master contract/subcontract database. 42 NERA Economic Consulting Table 5. County Distribution of WSDOT Contract Dollars County State Construction (%) Consulting (%) Combined (%) Asotin Benton Chelan Clallam Clark Columbia Cowlitz Douglas Franklin Grant Grays Harbor Island Jefferson King Kitsap Kittitas Klickitat Lewis Lincoln Mason Okanogan Pacific Pend Oreille Pierce Skagit Skamania Snohomish Spokane Stevens Thurston WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA WA 1.03 1.12 1.72 1.11 1.87 0.01 0.56 0.36 0.95 1.11 1.36 0.25 0.00 30.83 15.50 0.08 0.20 0.38 0.02 0.23 0.15 0.00 0.05 6.85 1.27 0.00 12.84 11.93 0.63 1.29 1.56 2.45 0.63 1.70 92.04 0.08 0.14 1.09 0.98 1.04 1.67 1.03 1.77 0.01 0.52 0.33 0.90 1.07 1.26 0.23 0.00 34.77 14.38 0.08 0.19 0.35 0.02 0.21 0.14 0.00 0.05 6.56 1.18 0.00 12.27 11.23 0.59 1.30 43 NERA Economic Consulting County State Construction (%) Consulting (%) Combined (%) Wahkiakum Walla Walla Whatcom Whitman Yakima TOTAL WA WA WA WA WA 0.00 0.02 2.07 0.01 4.22 100.0 0.00 0.02 0.30 0.01 100.0 1.94 0.01 3.91 100.0 Source: NERA calculations from WSDOT contracts databases. 44 NERA Economic Consulting Table 6. Total Businesses and Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 1611 1622 8711 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 7353 4212 1711 1781 4959 8743 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Engineering Services Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Public Relations Services Surveying Services TOTAL 726 36 2696 561 1071 2136 1383 302 2375 1890 1136 1709 34 71 150 370 2544 2634 12197 68 190 2542 2579 179 128 446 296 40449 39.21 15.22 6.53 5.38 5.34 4.83 3.69 2.64 2.18 1.89 1.74 1.60 1.59 1.30 1.14 1.13 0.90 0.83 0.61 0.45 0.36 0.34 0.27 0.25 0.23 0.19 0.18 39.21 54.43 60.95 66.34 71.68 76.51 80.20 82.84 85.02 86.90 88.64 90.24 91.83 93.12 94.27 95.40 96.30 97.13 97.74 98.19 98.54 98.88 99.16 99.41 99.63 99.82 100.00 45 NERA Economic Consulting Table 7. Construction Businesses and Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 1611 1622 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 8711 7353 4212 1711 1781 4959 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Engineering Services Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Surveying Services TOTAL 726 36 561 1071 2136 1383 302 2375 1890 1136 1709 34 71 150 370 2544 2634 12197 68 2696 190 2542 2579 179 128 296 40003 41.87 16.25 5.75 5.70 5.16 3.94 2.82 2.33 2.02 1.86 1.71 1.69 1.39 1.22 1.21 0.96 0.89 0.65 0.48 0.42 0.38 0.37 0.29 0.26 0.24 0.16 41.87 58.12 63.87 69.57 74.73 78.67 81.49 83.81 85.83 87.69 89.39 91.09 92.47 93.69 94.90 95.86 96.75 97.40 97.88 98.30 98.68 99.04 99.33 99.60 99.84 100.00 46 NERA Economic Consulting Table 8. Consulting Businesses and Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 8711 8743 8741 8748 8713 8742 Engineering Services Public Relations Services Management Services Business Consulting, n.e.c. Surveying Services Management Consulting Services TOTAL 2696 446 1032 5001 296 5180 14651 93.16 2.90 2.31 0.86 0.42 0.35 93.16 96.06 98.37 99.23 99.65 100.00 47 NERA Economic Consulting Table 9. Listed DBEs and Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 1611 1622 8711 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 7353 4212 1711 1781 4959 8743 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Engineering Services Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Public Relations Services Surveying Services TOTAL 65 5 247 44 83 177 96 22 241 136 96 75 3 5 16 38 359 236 2739 7 13 247 158 9 14 117 20 5268 39.21 15.22 6.53 5.38 5.34 4.83 3.69 2.64 2.18 1.89 1.74 1.60 1.59 1.30 1.14 1.13 0.90 0.83 0.61 0.45 0.36 0.34 0.27 0.25 0.23 0.19 0.18 39.21 54.43 60.95 66.34 71.68 76.51 80.20 82.84 85.02 86.90 88.64 90.24 91.83 93.12 94.27 95.40 96.30 97.13 97.74 98.19 98.54 98.88 99.16 99.41 99.63 99.82 100.00 48 NERA Economic Consulting Table 10. Listed Construction DBEs & Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 1611 1622 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 8711 7353 4212 1711 1781 4959 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Engineering Services Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Surveying Services TOTAL 65 5 44 83 177 96 22 241 136 96 75 3 5 16 38 359 236 2739 7 247 13 247 158 9 14 20 5151 41.87 16.25 5.75 5.70 5.16 3.94 2.82 2.33 2.02 1.86 1.71 1.69 1.39 1.22 1.21 0.96 0.89 0.65 0.48 0.42 0.38 0.37 0.29 0.26 0.24 0.16 41.87 58.12 63.87 69.57 74.73 78.67 81.49 83.81 85.83 87.69 89.39 91.09 92.47 93.69 94.90 95.86 96.75 97.40 97.88 98.30 98.68 99.04 99.33 99.60 99.84 100.00 49 NERA Economic Consulting Table 11. Listed Consulting DBEs & Industry Weight, by SIC Code SIC Code SIC Description Number of Establishments Industry Weight Industry Weight (Cumulative) 8711 8743 8741 8748 8713 8742 Engineering Services Public Relations Services Management Services Business Consulting, n.e.c. Surveying Services Management Consulting Services TOTAL 247 117 133 999 20 1094 2610 93.16 2.90 2.31 0.86 0.42 0.35 93.16 96.06 98.37 99.23 99.65 100.00 50 NERA Economic Consulting Table 12. Listed DBE Percentage & Industry Weight, by SIC Code SIC Code SIC Description Percentage Industry Weight Industry Weight (Cumulative) 1611 1622 8711 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 7353 4212 1711 1781 4959 8743 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Engineering Services Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Public Relations Services Surveying Services TOTAL 8.95 13.89 9.16 7.84 7.75 8.29 6.94 7.28 10.15 7.20 8.45 4.39 8.82 7.04 10.67 10.27 14.11 8.96 22.46 10.29 6.84 9.72 6.13 5.03 10.94 26.23 6.76 13.02 39.21 15.22 6.53 5.38 5.34 4.83 3.69 2.64 2.18 1.89 1.74 1.60 1.59 1.30 1.14 1.13 0.90 0.83 0.61 0.45 0.36 0.34 0.27 0.25 0.23 0.19 0.18 39.21 54.43 60.95 66.34 71.68 76.51 80.20 82.84 85.02 86.90 88.64 90.24 91.83 93.12 94.27 95.40 96.30 97.13 97.74 98.19 98.54 98.88 99.16 99.41 99.63 99.82 100.00 51 NERA Economic Consulting Table 13. Listed Construction DBE Percentage & Industry Weight, by SIC Code SIC Code SIC Description Percentage Industry Weight Industry Weight (Cumulative) 1611 1622 1629 1771 1731 1794 1541 1721 1542 4213 1751 2951 1442 1791 1623 782 1799 7389 1795 8711 7353 4212 1711 1781 4959 8713 Highway and Street Construction Bridge, Tunnel, and Elevated Highway Heavy Construction, n.e.c. Concrete Work Electrical Work Excavation Work Industrial Buildings and Warehouses Painting Nonresidential Construction, n.e.c. Trucking, Except Local Carpentry Work Paving Mixtures and Blocks Construction Sand and Gravel Structural Steel Erection Water, Sewer, and Utility Lines Lawn and Garden Services Special Trade Contractors, n.e.c. Business Services, n.e.c. Wrecking and Demolition Work Engineering Services Heavy Construction Equipment Rental and Leasing Local Trucking Without Storage Plumbing, Heating, and Air Conditioning Water Well Drilling Sanitary Services, n.e.c. Surveying Services TOTAL 8.95 13.89 7.84 7.75 8.29 6.94 7.28 10.15 7.20 8.45 4.39 8.82 7.04 10.67 10.27 14.11 8.96 22.46 10.29 9.16 6.84 9.72 6.13 5.03 10.94 6.76 12.88 41.87 16.25 5.75 5.70 5.16 3.94 2.82 2.33 2.02 1.86 1.71 1.69 1.39 1.22 1.21 0.96 0.89 0.65 0.48 0.42 0.38 0.37 0.29 0.26 0.24 0.16 41.87 58.12 63.87 69.57 74.73 78.67 81.49 83.81 85.83 87.69 89.39 91.09 92.47 93.69 94.90 95.86 96.75 97.40 97.88 98.30 98.68 99.04 99.33 99.60 99.84 100.00 52 NERA Economic Consulting Table 14. Listed Consulting DBE Percentage & Industry Weight, by SIC Code SIC Code SIC Description Percentage Industry Weight Industry Weight (Cumulative) 8711 8743 8741 8748 8713 8742 Engineering Services Public Relations Services Management Services Business Consulting, n.e.c. Surveying Services Management Consulting Services TOTAL 9.16 26.23 12.89 19.98 6.76 21.12 17.81 93.16 2.90 2.31 0.86 0.42 0.35 93.16 96.06 98.37 99.23 99.65 100.00 53 NERA Economic Consulting Table 15. Listed DBE Survey--Amount of Misclassification, by SIC Code Grouping52 Listed DBE By SIC Code Grouping SIC 16 SIC 15 SIC 17 SIC 87 SIC 42 Balance of SIC Codes All SIC Codes Misclassification (Percentage White Male) 26.9 21.2 18.9 19.1 24.1 15.8 21.2 Percentage Actually DBEowned 73.1 78.8 81.1 80.9 75.9 84.2 78.8 Number of Businesses Interviewed 67 52 53 89 54 38 353 Source: NERA telephone surveys conducted in February and March 2005. 52 SIC 16 - Heavy Construction, SIC 15 - Building Construction, SIC 17 - Special Trades Construction, SIC 87 - Professional Engineering and Other Services, SIC 42 - Trucking and Other Transportation. 54 NERA Economic Consulting Table 16. Listed DBE Survey--Amount of Misclassification, by Highway Region Misclassification Highway Region (Percentage White Male) Eastern North Central Northwest Olympic South Central Southwest Entire Region Source: See Table 15. Percentage Actually DBEowned 76.7 66.7 78.9 77.3 86.5 79.4 78.8 Number of Businesses Interviewed 30 15 171 66 37 34 353 23.3 33.3 21.1 22.7 13.5 20.6 21.2 55 NERA Economic Consulting Table 17. Listed DBE Survey--Amount of Misclassification, by Putative DBE Type Misclassific Putative Race/Sex ation (Percentage White Male) Black (either sex) Hispanic (either sex) Asian (either sex) Native American (either sex) Unspecified Minority (either sex) White Female All DBE Types Source: See Table 15. Misclassification (Percentage Other DBE Type) 5.2 9.1 3.9 4.0 77.8 8.2 8.3 Percentage Correctly Classified 89.5 75.0 84.6 87.8 0.0 62.2 70.5 Number of Businesses Interviewed 19 44 52 49 9 180 353 5.3 15.9 11.5 8.2 22.2 30.6 21.2 56 NERA Economic Consulting Table 18. Unclassified Businesses Survey --By SIC Code Grouping Listed DBE By SIC Code Grouping Stratum 1 Stratum 2 Stratum 3 Stratum 4 Stratum 5 Stratum 6 All SIC Codes 5.6 13.3 15.7 7.8 19.6 34.5 11.4 Percentage DBE Percentage Actually White Male-owned 94.4 86.7 84.3 92.2 80.4 65.5 88.6 Number of Businesses Interviewed 231 60 89 64 46 29 519 Source: NERA telephone surveys conducted in February and March 2005. 57 NERA Economic Consulting Table 19. Unclassified Businesses Survey --By Highway Region Percentage Highway Region Percentage DBE Actually White Male-owned Eastern North Central Northwest Olympic South Central Southwest Statewide Source: See Table 15. Number of Businesses Interviewed 45 20 206 141 43 64 519 11.1 10.0 12.6 12.1 9.3 7.8 11.4 88.9 90.0 87.4 87.9 90.7 92.2 88.6 58 NERA Economic Consulting Table 20. Unclassified Businesses Survey--By Race and Sex Number of Verified Race/Sex Businesses Interviewed White Male White Female Black Hispanic Asian Native American Statewide Source: See Table 18. Percentage of Total 88.6 6.4 0.8 1.2 1.7 1.4 100.0 460 33 4 6 9 7 519 59 NERA Economic Consulting Table 21. Calculation Summary--Overall Step / Calculation Number of Percentage Businesses of Total 40,449 5,268 3,931 3,939 3,955 7,474 4,471 11,412 7,592 100.00 13.02 9.72 9.74 9.78 18.48 11.05 28.21 18.77 All Businesses Listed DBEs Listed DBEs (effective number, with industry weights) Listed DBEs (effective number, corrected for misclassification) Listed DBEs (effective number, corrected for misclassif.; with industry weights) Unlisted DBEs (effective number, corrected for misclassification) Unlisted DBEs (effective number, corrected for misclassif.; with industry weights) All DBEs (final, unweighted) All DBEs (final, with industry weights) 60 NERA Economic Consulting Table 22. Calculation Summary--Construction Step / Calculation Number of Businesses 40,003 5,151 3,926 3,843 3,801 7,448 4,773 11,291 7,838 Percentage of Total 100.00 12.88 9.81 9.61 9.50 18.62 11.93 28.23 19.59 All Businesses Listed DBEs Listed DBEs (effective number, with industry weights) Listed DBEs (effective number, corrected for misclassification) Listed DBEs (effective number, corrected for misclassif.; with industry weights) Unlisted DBEs (effective number, corrected for misclassification) Unlisted DBEs (effective number, corrected for misclassif.; with industry weights) All DBEs (effective number, final, unweighted) All DBEs (effective number, final, with industry weights) 61 NERA Economic Consulting Table 23. Calculation Summary--Consulting Step / Calculation Number of Businesses 14,651 2,610 1,398 2,133 2,136 940 942 3,074 2,181 Percentage of Total 100.00 17.81 9.54 14.56 14.58 6.42 6.43 20.98 14.88 All Businesses Listed DBEs Listed DBEs (effective number, with industry weights) Listed DBEs (effective number, corrected for misclassification) Listed DBEs (effective number, corrected for misclassif.; with industry weights) Unlisted DBEs (effective number, corrected for misclassification) Unlisted DBEs (effective number, corrected for misclassif.; with industry weights) All DBEs (effective number, final, unweighted) All DBEs (effective number, final, with industry weights) 62 NERA Economic Consulting Table 24. Estimated DBE Availability for WSDOT Geographic Region Eastern North Central Northwest Olympic South Central Southwest Overall 17.37 15.42 19.95 17.65 21.71 16.46 Construction 18.21 15.94 21.21 18.47 22.41 16.59 Consulting 11.91 10.28 15.55 12.59 18.17 15.46 White Male White Female Black Hispanic Asian Native American MBE DBE 81.23 11.86 0.55 1.30 2.68 2.39 6.91 18.77 80.41 12.43 0.59 1.46 2.61 2.51 7.16 19.59 85.12 9.29 0.35 0.49 3.01 1.75 5.59 14.88 ENTIRE GEOGRAPHIC MARKET AREA 18.77 19.59 14.88 Source: (i) NERA calculations from master WSDOT contract/subcontract database; (ii) Dun & Bradstreet's MarketPlace; (iii) business directory information compiled by NERA; and (iv) NERA telephone surveys. 63 NERA Economic Consulting Table 25. Estimated DBE Utilization on WSDOT Construction Projects--Federally-Funded Only, Prime Contracts Only, Gross Contract Amount Contracts Type % N 586 25 4 1 3 3 11 36 624 % 96.71 2.15 0.40 0.02 0.51 0.14 1.07 3.22 100.00 $ 1,470,643,756 32,684,309 6,149,632 254,988 7,780,420 2,128,907 16,313,948 48,998,257 1,520,628,331 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 93.91 4.01 0.64 0.16 0.48 0.48 1.76 5.77 100.00 64 NERA Economic Consulting Table 26. Estimated DBE Utilization on WSDOT Construction Projects--Federally-Funded Only, Prime Contracts Only, NonSubcontracted Dollar Amounts Contracts Type % N 586 25 4 1 3 3 11 36 624 % 96.93 1.98 0.50 0.02 0.38 0.12 1.02 3.00 100.00 $ 934,339,651 19,118,874 4,838,814 150,733 3,659,171 1,155,349 9,804,068 28,922,942 963,981,488 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 93.91 4.01 0.64 0.16 0.48 0.48 1.76 5.77 100.00 65 NERA Economic Consulting Table 27. Estimated DBE Utilization on WSDOT Construction Projects--Federally-Funded Only, Prime and Subcontracts, First-Tier Only Contracts Type % N 4,266 974 61 109 85 175 428 1,384 5,622 % 86.31 8.22 1.99 1.30 1.33 1.51 6.12 14.32 100.00 $ 1,312,432,008 124,966,639 30,322,633 19,700,784 20,239,235 22,934,893 93,058,335 217,743,804 1,520,628,328 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 75.88 17.32 1.09 1.94 1.51 3.11 7.61 24.62 100.00 66 NERA Economic Consulting Table 28. Estimated DBE Utilization on WSDOT Construction Projects--Non-Federally-Funded Only, Prime Contracts Only, Gross Contract Amount Contracts Type % N 234 8 2 1 0 5 8 16 252 % 99.34 0.12 0.21 0.03 0.00 0.17 0.41 0.53 100.00 1,735,363 4,165,275 5,367,297 1,010,878,014 $ 1,004,227,834 1,202,021 2,091,603 338,309 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 92.86 3.17 0.79 0.40 0.00 1.98 3.17 6.35 100.00 67 NERA Economic Consulting Table 29. Estimated DBE Utilization on WSDOT Construction Projects--Non-Federally-Funded Only, Prime Contracts Only, Non-Subcontracted Dollar Amounts Contracts Type % N 234 8 2 1 0 5 8 16 252 % 99.35 0.13 0.18 0.04 0.00 0.18 0.40 0.53 100.00 1,529,073 3,363,102 4,494,565 842,982,672 $ 837,527,923 1,131,463 1,512,603 321,426 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 92.86 3.17 0.79 0.40 0.00 1.98 3.17 6.35 100.00 68 NERA Economic Consulting Table 30. Estimated DBE Utilization on WSDOT Construction Projects--Non-Federally-Funded Only, Prime and Subcontracts, First-Tier Only Contracts Type % N 1,350 242 13 30 20 43 104 337 1,693 % 96.98 1.53 0.33 0.36 0.10 0.67 1.45 2.97 100.00 $ 980,387,520 15,447,446 3,325,453 3,621,541 999,372 6,740,693 14,671,032 29,991,946 1,010,878,015 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 79.74 14.29 0.77 1.77 1.18 2.54 6.14 19.91 100.00 69 NERA Economic Consulting Table 31. Estimated DBE Utilization on WSDOT Consulting Projects--Federally-Funded Only, Prime Contracts Only, Gross Contract Amount Contracts Type % N 6 4 0 2 0 1 3 7 89 % 3.10 0.47 0.00 2.23 0.00 0.30 2.53 3.00 100.00 324,555 2,707,316 3,213,349 107,025,548 2,382,761 $ 3,312,947 506,033 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 6.74 4.49 0.00 2.25 0.00 1.12 3.37 7.87 100.00 70 NERA Economic Consulting Table 32. Estimated DBE Utilization on WSDOT Consulting Projects--Federally-Funded Only, Prime Contracts Only, NonSubcontracted Dollar Amounts Contracts Type % N 6 4 0 2 0 1 3 7 89 % 5.18 0.79 0.00 3.66 0.00 0.51 4.17 4.96 100.00 324,555 2,669,356 3,175,389 64,009,932 2,344,801 $ 3,312,947 506,033 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 6.74 4.49 0.00 2.25 0.00 1.12 3.37 7.87 100.00 71 NERA Economic Consulting Table 33. Estimated DBE Utilization on WSDOT Consulting Projects--Federally-Funded Only, Prime and Subcontracts, First-Tier Only Contracts Type % N 11 26 2 6 14 2 24 50 206 % 3.92 6.33 0.05 2.60 1.12 0.34 4.11 10.44 100.00 $ 4,196,987 6,775,918 54,479 2,783,686 1,199,543 358,903 4,396,611 11,172,529 107,025,548 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 5.34 12.62 0.97 2.91 6.80 0.97 11.65 24.27 100.00 72 NERA Economic Consulting Table 34. Estimated DBE Utilization on WSDOT Consulting Projects--Non-Federally-Funded Only, Prime Contracts Only, Gross Contract Amount Contracts Type % N 23 9 0 4 4 0 8 17 240 % 8.16 3.20 0.00 0.77 0.37 0.00 1.14 4.34 100.00 1,785,420 6,790,710 156,427,228 1,203,935 581,485 $ 12,769,411 5,005,290 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 9.58 3.75 0.00 1.67 1.67 0.00 3.33 7.08 100.00 73 NERA Economic Consulting Table 35. Estimated DBE Utilization on WSDOT Consulting Projects--Non-Federally-Funded Only, Prime Contracts Only, Non-Subcontracted Dollar Amounts Contracts Type % N 23 9 0 4 4 0 8 17 240 % 10.28 4.03 0.00 0.97 0.44 0.00 1.41 5.44 100.00 1,739,852 6,698,009 123,104,599 1,196,497 543,355 $ 12,660,373 4,958,157 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 9.58 3.75 0.00 1.67 1.67 0.00 3.33 7.08 100.00 74 NERA Economic Consulting Table 36. Estimated DBE Utilization on WSDOT Consulting Projects--Non-Federally-Funded Only, Prime and Subcontracts, First-Tier Only Contracts Type % N 34 67 5 14 35 0 54 121 510 % 8.89 5.69 0.28 1.38 3.31 0.00 4.97 10.66 100.00 7,772,898 16,678,233 156,427,229 $ 13,905,701 8,905,335 439,953 2,158,310 5,174,635 Contract Dollars White Male White Female Black Hispanic Asian/Pacific Native American All MBE All DBE TOTAL 6.67 13.14 0.98 2.75 6.86 0.00 10.59 23.73 100.00 75 NERA Economic Consulting Table 37. Annual Wage Earnings Regressions, All Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Yes Yes Yes 3848837 .436 18480 (1) -0.304 (197.61) -0.216 (139.09) -0.292 (139.06) -0.327 (70.23) -0.281 (89.02) -0.357 (400.16) 0.177 (680.45) -0.002 (588.53) 0.213 (17.80) Specification (2) -0.305 (197.36) -0.217 (138.95) -0.293 (137.52) -0.329 (69.08) -0.283 (88.14) -0.357 (396.29) 0.177 (680.42) -0.002 (588.51) 0.197 (15.87) 0.103 (4.62) 0.075 (5.87) 0.056 (4.03) 0.068 (2.46) 0.084 (4.12) 0.000 (.01) Yes Yes Yes 3848837 .436 17816 (3) -0.305 (197.51) -0.217 (139.06) -0.293 (137.58) -0.329 (69.09) -0.283 (88.15) -0.357 (400.18) 0.177 (680.42) -0.002 (588.51) 0.197 (16.48) 0.103 (4.67) 0.075 (6.06) 0.056 (4.13) 0.068 (2.47) 0.084 (4.18) Yes Yes Yes 3848837 .436 18032 Source: NERA calculations from the 2000 Decennial Census Five Percent Public Use Microdata Samples. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. 76 NERA Economic Consulting Table 38. Annual Wage Earnings Regressions, All Industries, 19791991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (13 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N R2 F Yes Yes Yes Yes 1868379 .504 16243 (1) -0.220 (205.27) -0.167 (122.92) -0.194 (109.06) -0.238 (370.55) 0.057 (351.86) -0.001 (286.2) -0.061 (14.01) Specification (2) -0.220 (204.87) -0.167 (122.76) -0.194 (107.96) -0.238 (368.36) 0.057 (351.86) -0.001 (286.21) -0.060 (11.82) 0.019 (.99) 0.045 (2.13) 0.006 (.44) -0.006 (1.07) Yes Yes Yes Yes 1868379 .504 15706 (3) -0.220 (205.27) -0.167 (122.81) -0.194 (109.07) -0.238 (370.55) 0.057 (351.86) -0.001 (286.20) -0.062 (14.15) 0.047 (2.24) Yes Yes Yes Yes 1868379 .504 16105 Source: NERA calculations from the Merged Outgoing Rotation Groups of the 19791991 Current Population Survey microdata samples. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 77 NERA Economic Consulting Table 39. Annual Wage Earnings Regressions, All Industries, 19922002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N R2 F Yes Yes Yes Yes 933024 .467 6372 (1) -0.214 (129.51) -0.206 (118.35) -0.194 (78.96) -0.171 (38.05) -0.178 (174.59) 0.053 (202.35) -0.001 (166.92) -0.070 (11.05) Specification (2) -0.214 (129.35) -0.206 (118.14) -0.195 (78.45) -0.172 (37.93) -0.178 (173.5) 0.053 (202.35) -0.001 (166.92) -0.077 (10.21) 0.059 (2.16) 0.039 (1.93) 0.050 (2.73) 0.068 (1.89) 0.002 (.27) Yes Yes Yes Yes 933024 .467 6133 (3) -0.214 (129.4) -0.206 (118.38) -0.195 (78.45) -0.171 (38.05) -0.178 (174.59) 0.053 (202.35) -0.001 (166.92) -0.074 (11.44) 0.055 (2.05) 0.046 (2.59) Yes Yes Yes Yes 933024 .467 6274 Source: NERA calculations from the Merged Outgoing Rotation Groups of the 19922002 Current Population Survey microdata samples. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 78 NERA Economic Consulting Table 40. Annual Wage Earnings Regressions, Construction and Related Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Source: See Table 37. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64 employed in the construction or construction-related professional services industries; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. Yes Yes Yes 307414 .268 1503 (1) -0.334 (52.33) -0.158 (31.74) -0.195 (17.87) -0.296 (22.41) -0.216 (18.73) -0.395 (103.90) 0.157 (174.96) -0.002 (149.34) 0.260 (7.22) Specification (2) -0.334 (52.26) -0.159 (31.72) -0.197 (17.75) -0.299 (22.18) -0.222 (19) -0.394 (102.19) 0.157 (174.97) -0.002 (149.35) 0.251 (6.92) 0.126 (1.35) 0.060 (1.34) 0.057 (.84) 0.102 (1.29) 0.233 (3.17) -0.041 (1.45) Yes Yes Yes 307414 .268 1392 (3) -0.334 (52.36) -0.158 (31.76) -0.195 (17.89) -0.296 (22.41) -0.222 (18.98) -0.395 (103.89) 0.157 (174.96) -0.002 (149.34) 0.254 (7.08) 0.231 (3.15) Yes Yes Yes 307414 .268 1484 79 NERA Economic Consulting Table 41. Annual Wage Earnings Regressions, Construction and Related Industries, 1979-1991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (13 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N R2 F Source: See Table 38. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64 employed in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 123230 .399 1169 (1) -0.213 (32.07) -0.139 (19.87) -0.098 (8.85) -0.287 (61.23) 0.070 (72.46) -0.001 (57.41) -0.034 (1.09) Specification (2) -0.213 (31.94) -0.139 (19.75) -0.097 (8.81) -0.287 (61.22) 0.070 (72.47) -0.001 (57.41) -0.039 (0.89) -0.267 (1.99) -0.006 (0.08) -0.057 (0.29) 0.136 (1.46) Yes Yes Yes Yes 123230 .399 1105 (3) -0.213 (31.93) -0.139 (19.89) -0.098 (8.85) -0.287 (61.24) 0.070 (72.47) -0.001 (57.42) -0.020 (0.63) -0.282 (2.16) Yes Yes Yes Yes 123230 .399 1105 80 NERA Economic Consulting Table 42. Annual Wage Earnings Regressions, Construction and Related Industries, 1992-2002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N R2 F Source: See Table 39. Notes: (1) Universe is all private sector prime age wage and salary workers between age 16 and 64 employed in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual wages between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 60581 .373 439 (1) -0.196 (25.63) -0.175 (29.57) -0.116 (9.05) -0.103 (7.2) -0.245 (48.99) 0.062 (61.08) -0.001 (47.95) 0.049 (1.8) Specification (2) -0.196 (25.56) -0.176 (29.63) -0.116 (8.84) -0.104 (7.26) -0.246 (48.67) 0.062 (61.08) -0.001 (47.95) 0.037 (1.33) -0.019 (.17) 0.144 (1.91) -0.015 (.21) 0.108 (.92) 0.042 (1) Yes Yes Yes Yes 60581 .373 413 (3) -0.196 (25.63) -0.175 (29.57) -0.116 (9.05) -0.103 (7.2) -0.245 (48.99) 0.062 (61.08) -0.001 (47.95) 0.049 (1.8) Yes Yes Yes Yes 60581 .373 439 81 NERA Economic Consulting Table 43. Annual Business Owner Earnings Regressions, All Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Yes Yes Yes 401629 .166 497 (1) -0.300 (26.46) -0.190 (18.84) -0.041 (2.86) -0.384 (14.84) -0.273 (15.1) -0.440 (90.29) 0.164 (98.38) -0.002 (88.4) -0.110 (2.18) Specification (2) -0.300 (26.35) -0.191 (18.79) -0.043 (2.93) -0.384 (14.55) -0.278 (15.12) -0.441 (89.6) 0.164 (98.38) -0.002 (88.4) -0.135 (2.61) -0.066 (.45) 0.033 (.38) 0.060 (.71) 0.024 (.15) 0.199 (1.61) 0.061 (1.73) Yes Yes Yes 401629 .166 482 (3) -0.300 (26.46) -0.190 (18.84) -0.041 (2.86) -0.384 (14.84) -0.273 (15.1) -0.440 (90.29) 0.164 (98.38) -0.002 (88.4) -0.110 (2.18) Yes Yes Yes 401629 .166 497 Source: NERA calculations from the 2000 Decennial Census Five Percent Public Use Microdata Samples. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, tstatistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. 82 NERA Economic Consulting Table 44. Annual Business Owner Earnings Regressions, All Industries, 1979-1991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (13 categories) Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Yes Yes Yes Yes 82094 .177 153.23 (1) -0.500 (15.64) -0.278 (9.46) -0.328 (8.29) -0.729 (68.07) 0.205 (41.42) -0.002 (36.5) -0.238 (2.53) Specification (2) -0.502 (15.69) -0.280 (9.46) -0.329 (8.19) -0.731 (67.87) 0.205 (41.41) -0.002 (36.49) -0.326 (3.33) 0.623 (.77) 0.072 (.21) 0.080 (.27) 0.384 (2.61) Yes Yes Yes Yes 82094 .177 148.14 (3) -0.501 (15.67) -0.279 (9.48) -0.327 (8.26) -0.731 (67.87) 0.205 (41.41) -0.002 (36.48) -0.316 (3.29) 0.364 (2.56) Yes Yes Yes Yes 82094 .177 151.97 Source: NERA calculations from the Annual Demographic (March) File of the 19791991 Current Population Survey microdata samples. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 83 NERA Economic Consulting Table 45. Annual Business Owner Earnings Regressions, All Industries, 1992-2002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Yes Yes Yes Yes 55639 .128 64.4 (1) -0.591 (14.85) -0.390 (9.8) -0.221 (3.41) -0.511 (5.47) -0.617 (31.34) 0.230 (27.27) -0.002 (23.8) 0.237 (1.29) Specification (2) (3) -0.589 -0.591 (14.74) (14.85) -0.394 -0.390 (9.89) (9.8) -0.214 -0.221 (3.25) (3.41) -0.504 -0.511 (5.29) (5.47) -0.617 -0.617 (31.08) (31.34) 0.230 0.230 (27.28) (27.27) -0.002 -0.002 (23.81) (23.8) 0.259 0.237 (1.2) (1.29) -0.517 (.94) 1.360 (1.64) -0.284 (.75) -0.412 (.66) -0.028 (.14) Yes Yes Yes Yes Yes Yes Yes Yes 55639 .129 62.00 55639 .128 64.4 Source: NERA calculations from the Annual Demographic (March) File of the 19922002 Current Population Survey microdata samples. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 84 NERA Economic Consulting Table 46. Business Owner Earnings Regressions, Construction and Related Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Source: See Table 43. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64 in the construction or construction-related professional services industries; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, tstatistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. Yes Yes Yes 64853 .054 49 (1) -0.338 (12.11) -0.147 (6.89) -0.069 (1.47) -0.354 (7.01) -0.149 (3.41) -0.505 (30.56) 0.136 (36.01) -0.001 (33.71) -0.237 (1.46) Specification (2) -0.336 (11.97) -0.147 (6.84) -0.071 (1.51) -0.357 (6.98) -0.146 (3.3) -0.506 (30.1) 0.136 (36.01) -0.001 (33.71) -0.235 (1.44) -0.573 (1.65) -0.022 (.11) 0.123 (.36) 0.175 (.46) -0.097 (.37) 0.020 (.16) Yes Yes Yes 64853 .054 46 (3) -0.338 (12.11) -0.147 (6.89) -0.069 (1.47) -0.354 (7.01) -0.149 (3.41) -0.505 (30.56) 0.136 (36.01) -0.001 (33.71) -0.237 (1.46) Yes Yes Yes 64853 .054 49 85 NERA Economic Consulting Table 47. Business Owner Earnings Regressions, Construction and Related Industries, 1979-1991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (13 categories) Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Source: See Table 44. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64 in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 12577 .077 14.99 (1) -0.428 (5.73) -0.252 (3.96) -0.208 (1.79) -0.835 (21.63) 0.179 (16.58) -0.002 (15.29) -0.260 (1.13) Specification (2) (3) -0.428 -0.428 (5.72) (5.74) -0.258 -0.252 (4.05) (3.96) -0.175 -0.175 (1.47) (1.47) -0.839 -0.839 (21.77) (21.76) 0.178 0.178 (16.56) (16.57) -0.002 -0.002 (15.29) (15.3) -0.286 -0.267 (1.25) (1.16) -0.108 (.09) 1.256 (1.1) -0.981 -0.982 (3.12) (3.14) 4.432 4.301 (2.28) (2.25) Yes Yes Yes Yes Yes Yes Yes Yes 12577 .079 14.42 12577 .079 14.80 86 NERA Economic Consulting Table 48. Business Owner Earnings Regressions, Construction and Related Industries, 1992-2002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (16 categories) Geography (51 categories) Industry (88 categories) N R2 F Source: See Table 45. Notes: (1) Universe is all persons in the private sector with positive business income between age 16 and 64 in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number is the percentage difference in annual business earnings between a given group and white men; (3) Number in parentheses is the absolute value of the associated t-statistic. Using a two-tailed test, t-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 8446 .064 6.97 0.184 (.22) Yes Yes Yes Yes 8446 .064 6.73 -0.047 (.03) 1.840 (.79) (1) -0.323 (2.4) -0.145 (1.38) -0.180 (.84) -0.208 (.76) -0.839 (15.73) 0.190 (8.71) -0.002 (7.89) 0.560 (1.11) Specification (2) -0.323 (2.4) -0.145 (1.38) -0.207 (.97) -0.208 (.76) -0.840 (15.56) 0.190 (8.71) -0.002 (7.9) 0.479 (.94) (3) -0.323 (2.4) -0.145 (1.38) -0.180 (.84) -0.208 (.76) -0.839 (15.73) 0.190 (8.71) -0.002 (7.89) 0.560 (1.11) Yes Yes Yes Yes 8446 .064 6.97 87 NERA Economic Consulting Table 49. Self-Employment Rates in 2000 for Selected Race and Sex Groups: All Industries; United States and the State of Washington Race/Sex Black Hispanic Asian Native American Multiple Races White female White male U.S. 5.1 7.3 10.1 8.4 9.2 8.2 13.6 State of Washington 5.7 5.9 9.3 8.0 8.3 10.4 13.3 Source: NERA calculations from the 2000 Decennial Census Five Percent Public Use Microdata Samples. 88 NERA Economic Consulting Table 50. Self-Employment Rates in 2000 for Selected Race and Sex Groups: Construction and Related Industries; United States and the State of Washington Race/Sex Black Hispanic Asian Native American Multiple Races White female White male U.S. (%) 14.9 12.9 16.7 16.7 20.4 14.7 25.0 Washington 5.5 10.5 13.4 13.3 10.3 14.5 21.9 Source: NERA calculations from the 2000 Decennial Census Five Percent Public Use Microdata Samples. 89 NERA Economic Consulting Table 51. Business Formation Regressions, All Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (25 categories) N Pseudo R2 Chi2 Log Likelihood Yes Yes Yes 4406525 0.162 480000 -1255764 (1) -0.047 (104.87) -0.036 (85.08) -0.016 (26.13) -0.033 (26.22) -0.018 (19.75) -0.030 (105.61) 0.011 (152.62) -0.000 (108.22) 0.025 (7.07) Specification (2) -0.047 (104.88) -0.036 (84.19) -0.016 (26.05) -0.033 (25.48) -0.018 (19.65) -0.031 (106.63) 0.011 (152.59) -0.000 (108.2) 0.014 (3.91) 0.013 (1.83) -0.019 (5.49) 0.011 (2.86) 0.001 (0.12) 0.012 (2.09) 0.030 (15.12) Yes Yes Yes 4406525 0.162 480000 -1255610 (3) -0.047 (105.10) -0.036 (84.18) -0.016 (26.04) -0.033 (26.07) -0.018 (19.65) -0.031 (106.62) 0.011 (152.59) -0.000 (108.20) 0.014 (4.04) -0.019 (5.59) 0.011 (2.78) 0.012 (2.03) 0.029 (15.09) Yes Yes Yes 4406525 0.162 480000 -1255612 Source: NERA calculations from the 2000 Decennial Census Five Percent Public Use Microdata Samples. Notes: (1) Universe is all private sector prime age labor force participants between age 16 and 64; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, z-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. 90 NERA Economic Consulting Table 52. Business Formation Regressions, All Industries, 1979-1991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (6 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N Pseudo R2 Chi2 Log Likelihood Yes Yes Yes Yes 2684590 .245 4.4e+05 -671453 (1) -0.037 (93.66) -0.028 (58.68) -0.016 (25.92) -0.027 (100.93) 0.011 (178.81) -0.000 (139.92) 0.023 (12.85) Specification (2) -0.037 (93.69) -0.028 (58.66) -0.016 (25.97) -0.027 (100.97) 0.011 (.178.81) -0.000 (139.91) 0.016 (7.80) 0.026 (2.70) 0.008 (1.00) 0.014 (2.70) 0.014 (6.38) Yes Yes Yes Yes 2684590 .245 4.4e+05 -671430 (3) -0.037 (93.68) -0.028 (58.76) -0.016 (25.97) -0.027 (100.97) 0.011 (.178.81) -0.000 (139.91) 0.016 (7.80) 0.025 (2.67) 0.014 (2.66) 0.014 (6.31) Yes Yes Yes Yes 2684590 .245 4.4e+05 -671430 Source: NERA calculations from the Merged Outgoing Rotation Groups of the 19791991 Current Population Survey microdata samples. Notes: (1) Universe is all private sector prime age labor force participants between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, z-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 91 NERA Economic Consulting Table 53. Business Formation Regressions, All Industries, 1992-2002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N Pseudo R2 Chi2 Log Likelihood Yes Yes Yes Yes 1924167 .215 3.1e+05 -568248 (1) -0.048 (78.37) -0.041 (61.79) -0.015 (16.51) -0.030 (19.24) -0.026 (62.43) 0.013 (125.43) -0.000 (89.59) 0.010 (4.07) Specification (2) -0.048 (78.41) -0.041 (61.50) -0.016 (16.65) -0.030 (18.98) -0.026 (62.66) 0.013 (125.44) -0.000 (89.60) 0.002 (0.69) 0.031 (2.24) -0.023 (2.69) 0.020 (2.56) -0.005 (0.38) 0.020 (5.72) Yes Yes Yes Yes 1924167 .215 3.1e+05 -568222 (3) -0.048 (78.41) -0.041 (61.50) -0.016 (16.65) -0.030 (19.22) -0.026 (62.67) 0.013 (125.44) -0.000 (89.60) 0.002 (0.65) 0.032 (2.25) -0.023 (2.68) 0.020 (2.58) 0.020 (5.78) Yes Yes Yes Yes 1924167 .215 3.1e+05 -568222 Source: NERA calculations from the Merged Outgoing Rotation Groups of the 19922002 Current Population. Notes: (1) Universe is all private sector prime age labor force participants between age 16 and 64; observations with imputed earnings are excluded where identified; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, z-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. 92 NERA Economic Consulting Table 54. Business Formation Regressions, Construction and Related Industries, 2000 Independent Variables Black Hispanic Asian/Pacific Islanders Native American Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington* Asian/Pacific Islanders Washington* Native American Washington*Other Race Washington*White Female Education (16 categories) Geography (51 categories) Industry (25 categories) N Pseudo R2 Chi2 Log Likelihood Source: See Table 51. Notes: (1) Universe is all private sector prime age labor force participants in the construction sector between age 16 and 64; observations with imputed values to the dependent variable and all independent variables are excluded; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, zstatistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes persons identifying themselves as belonging in more than one racial category; (5) Geography is defined based on place of residence. Specification (2) -0.096 (30.87) -0.076 (32.13) -0.056 (10.48) -0.076 (11.55) -0.029 (5.25) -0.087 (41.36) 0.026 (63.86) -0.000 (46.81) -0.026 (1.29) - 0.103 (2.25) 0.010 (0.40) 0.013 (0.37) -0.004 (0.10) -0.024 (0.74) 0.045 (3.05) Yes Yes Yes Yes Yes Yes 376898 376898 .075 .075 30026 30042 -184677 -184669 (1) -0.097 (31.11) -0.076 (32.23) -0.056 (10.58) -0.076 (11.82) -0.030 (5.47) -0.086 (41.45) 0.026 (63.86) -0.000 (46.81) -0.023 (1.13) (3) -0.096 (30.88) -0.076 (32.24) -0.056 (10.58) -0.076 (11.81) -0.030 (5.47) -0.087 (41.36) 0.026 (63.86) -0.000 (46.81) -0.026 (1.29) - 0.103 (2.25) 0.045 (3.07) Yes Yes Yes 376898 .075 30026 -184670 93 NERA Economic Consulting Table 55. Business Formation Regressions, Construction and Related Industries, 1979-1991 Independent Variables Black Hispanic Other Race White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Other Race Washington*White Female Time (6 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N Pseudo R2 Chi2 Log Likelihood Source: See Table 52. Notes: (1) Universe is all private sector prime age labor force participants between age 16 and 64 in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, z-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 209444 .082 16816 -93584 (1) -0.085 (25.12) -0.065 (16.79) -0.095 (18.24) -0.099 (36.87) 0.028 (61.25) -0.000 (49.49) -0.014 (1.07) Specification (2) -0.085 (25.11) -0.065 (16.83) -0.095 (18.13) -0.099 (36.55) 0.028 (61.25) -0.000 (49.49) -0.015 (1.20) 0.047 (0.71) 0.082 (1.30) 0.026 (0.52) -0.001 (0.06) Yes Yes Yes Yes 209444 .083 16819 -93583 (3) -0.085 (25.12) -0.065 (16.79) -0.095 (18.24) -0.099 (36.87) 0.028 (61.25) -0.000 (49.49) -0.014 (1.07) Yes Yes Yes Yes 209444 .082 16816 -93584 94 NERA Economic Consulting Table 56. Business Formation Regressions, Construction and Related Industries, 1992-2002 Independent Variables Black Hispanic Asian Native American White Female Age Age2 Washington Washington*Black Washington*Hispanic Washington*Asian Washington*Native American Washington*White Female Time (11 categories) Education (continuous) Geography (51 categories) Industry (49 categories) N Pseudo R2 Chi2 Log Likelihood Source: See Table 53. Notes: (1) Universe is all private sector prime age labor force participants between age 16 and 64 in the construction or construction-related professional services industries; observations with imputed earnings are excluded where identified; (2) Reported number represents the percentage point probability difference in business ownership rates between a given group and white men, evaluated at the mean business ownership rate for the estimation sample; (3) Number in parentheses is the absolute value of the associated z-statistic. Using a two-tailed test, z-statistics greater than 1.67 (1.99) (2.64) are statistically significant at a 90 (95) (99) percent confidence level; (4) "Other Race" includes Hispanics, Asian/Pacific Islanders, and American Indians/Alaska Natives; (5) Geography is defined based on place of residence. Yes Yes Yes Yes 153805 .090 15294 -77525 (1) -0.110 (23.82) -0.091 (21.01) -0.075 (8.94) -0.089 (10.1) -0.048 (13.72) 0.033 (48.78) -0.000 (36.89) -0.012 (.69) Specification (2) -0.110 (23.84) -0.091 (20.91) -0.075 (8.76) -0.089 (9.99) -0.049 (13.84) 0.033 (48.79) -0.000 (36.89) -0.018 (0.96) 0.101 (1.06) -0.044 (0.75) -0.034 (0.54) -0.008 (0.10) 0.051 (1.82) Yes Yes Yes Yes 153805 .090 15300 -77523 (3) -0.110 (23.82) -0.091 (21.01) -0.075 (8.94) -0.089 (10.1) -0.048 (13.72) 0.033 (48.78) -0.000 (36.89) -0.012 (.69) Yes Yes Yes Yes 153805 .090 15294 -77525 95 NERA Economic Consulting Table 57. Actual and Potential Business Formation Rates-- Washington State Construction and Consulting Markets Race/Sex Business Formation Rate (%) (1) Potential Business Formation Rate (%) (2) 25.5 18.1 19.0 20.9 13.3 18.7 20.7 Disparity Ratio (3) 0.22 0.58 0.71 0.64 0.77 0.78 0.70 Black Hispanic Asian/Pacific Islander American Indian/Alaska Native Multiple races reported White female All minority and female 5.5 10.5 13.4 13.3 10.3 14.5 14.4 Notes: Figures in column (1) are average self-employment rates weighted using PUMS population-based person weights. Figures in column (2) are derived from combining the figure in column (1) with the corresponding result from Table 54. Column (3) is simply column (1) divided by column (2). Source: 2000: Five Percent PUMS. See Table 54. 96 NERA Economic Consulting Table 58. Comparison of Baseline to Adjusted DBE Availability for WSDOT Contracting Area Baseline DBE Availability (%) 19.59 14.88 18.77 Adjusted DBE Availability (%) 28.31 24.32 28.12 Construction Consulting TOTAL - All FEDERAL-AID PROJECTS Source: (1) WSDOT contract databases; (2) Dun & Bradstreet's MarketPlace; (3) business directory information compiled by NERA; (4) NERA telephone surveys; and (5) the Five Percent 2000 PUMS. 97 NERA Economic Consulting VIII. CONCLUSION In this study, NERA estimated the availability of minority-owned and womanowned businesses in WSDOT's contracting markets. This involved identifying the relevant markets for federally-assisted WSDOT contracting--that is, the main industries and localities where WSDOT spends its dollars. In consultation with WSDOT, NERA identified 26 distinct four-digit SIC codes in construction and 6 in consulting that account for virtually all contract, subcontract and supplier spending on WSDOT projects. We found that from FFY 1999 and FFY 2003, 94 percent of WSDOT's spending was with businesses located in the State of Washington, compared with 93 percent in consulting, and 94 percent overall. A further challenge was to count businesses in the relevant markets and determine the proportion that was owned by minorities and women. To count the number of businesses, we used Dun & Bradstreet's MarketPlace database to determine the total number operating in the relevant geographic and product markets. MarketPlace does not adequately identify all businesses owned by minorities and women however. NERA took a number of steps to overcome this limitation. First, we completed an intensive regional search for information on minority-owned and woman-owned businesses in and surrounding the Washington area. Second, we conducted a survey to check whether the ownership status of these businesses was correct--some firms classified as DBEs were in fact not minority-owned and vice versa. We found that of the firms that were listed as DBEs, more than one-in-five, on average, were wrongly classified and were actually owned by White males. Similarly, a large number of businesses in the MarketPlace database did not have the race or gender of their owners identified. Most, but not all, of these firms are likely to be White male owned. To test and quantify this, we conducted a second survey and found that on average 11.4 percent of these initially unclassified businesses were actually owned by women and/or minorities. Once the relevant product markets were established and we had an accurate estimate of the ownership status of the businesses in the database, we estimated final baseline DBE availability. Our final baseline estimates are 19.59 percent in construction, 14.88 percent in consulting, and 18.77 percent overall. Finally, Step 2 adjustments were estimated using data from the 2000 Five Percent PUMS to take account of the fact that the 98 NERA Economic Consulting baseline numbers are lower than what would be expected in a race and sex neutral marketplace. Step 2 adjusted availability estimates are 28.31 percent for construction, 24.32 percent for consulting, and 28.12 percent overall. 99 NERA Economic Consulting IX. REFERENCES [1] Z. Acs and D. Evans, "The determinants of variations in self-employment rates across countries and over time," working paper (1994). A. Alba-Ramirez, "Self-employment in the midst of unemployment; the case of Spain and the United States," Applied Economics 26 (1994): 189-204. A. B. Arai, "The road not taken: The transition from unemployment to self-employment in Canada, 1961-1994," Canadian Journal of Sociology 22 (Summer 1997): 365-382. R. L. Aronson, Self-employment. ILR Press, Ithaca, NY: ILR Press, 1991. T. Bates, "The changing nature of minority business: A comparative analysis of Asian, Nonminority, and Black-Owned businesses," The Review of Black Political Economy (1989): 25-42. I. Bernhardt, "Comparative advantage in self-employment and paid work," Canadian Journal of Economics (May 1994): 273-289. J. Black, D. De Meza, and D. Jeffreys, "House prices, the supply of collateral, and the enterprise economy," Economic Journal 106 (January 1996): 60-75. D. G. Blanchflower, "Self-Employment in OECD countries," Labour Economics 7 (September 2000): 471-505. D. G. Blanchflower, "Self-employment: More may not be better," working paper, National Bureau of Economic Research w10286, February 2004. D. G. Blanchflower and B. Meyer, "A longitudinal analysis of the young self-employed in Australia and the United States," Small Business Economics 6 (1994): 1-20. D. G. Blanchflower and A. J. Oswald, "Self-employment and the enterprise culture," British Social Attitudes: the 1990 Report. Edited by R. Jowell, and S. Witherspoon. Aldershot: Gower Press, 1990. D. G. Blanchflower and A. J. Oswald, "What makes an entrepreneur?" Journal of Labor Economics 16 no. 1 (January 1998): 26-60. D. G. Blanchflower and A. J. Oswald, The Wage Curve. Cambridge: MIT Press, 1994. D. G. Blanchflower, A. J. Oswald, and A. Stutzer, "Latent entrepreneurship across nations," European Economic Review 45 no.4-6 (May 2001): 680-691. D. Blau, "A time-series analysis of self-employment in the United States," Journal of Political Economy 95 (1987): 445-467. [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] 100 NERA Economic Consulting [16] D. Bogenhold and U. Staber, "The decline and rise of self-employment," Employment and Society 5 (1991): 223-239. G. J. Borjas and S. Bronars, "Consumer discrimination and self-employment," Journal of Political Economy 97 (1989): 581-605. N. Brossard, R. Chami, and G. Hess, "(Why) do self-employed parents have more children?" working paper (September 2003). S. Coate and S. Tennyson, "Labor market discrimination, imperfect information and selfemployment," Oxford Economic Papers 44 (1992): 272-288. M. Cowling and P. Mitchell, "The evolution of UK self-employment: A study of government policy and the role of the macroeconomy," Manchester School of Economic and Social Studies 65 no.4 (September 1997): 427-442. G. DeWit and F. A. Van Winden, "An empirical analysis of self-employment in the Netherlands," Economics Letters 32 (1990): 97-100. D. Evans and B. Jovanovic, "An estimated model of entrepreneurial choice under liquidity constraints," Journal of Political Economy 97 (1989): 808-827. D. Evans and L. Leighton, "Some empirical aspects of entrepreneurship," American Economic Review 79 (1989): 519-535. R. W. Fairlie, "The absence of the Black owned business: An analysis of the dynamics of selfemployment," Journal of Labor Economics 17 no.1 (1999): 80-108. R. W. Fairlie and B. D. Meyer, "Ethnic and racial self-employment differences and possible explanations," Journal of Human Resources 31 no. 4 (1996): 757-793. R. W. Fairlie and B. D. Meyer, "Does immigration hurt Black self-employment?" Help or Hindrance? The Economic Implications of Immigration for Blacks. Edited by D. S. Hamermesh and F. D. Bean. New York: Russell Sage Foundation, 1998 R. W. Fairlie and B. D. Meyer, "Trends in self-employment among white and black men during the twentieth century," Journal of Human Resources XXXV no.4 (2000): 643-669. R. W. Fairlie and B. D. Meyer, "The effect of immigration on native self-employment," forthcoming in Journal of Labor Economics (2003). A. Foti and M. Vivarelli, "An econometric test of the self-employment model - the case of Italy," Small Business Economics 6 no.2 (April 1994): 81-93. Victor Fuchs, "Self-employment and labor force participation of older males," Journal of Human Resources 17 (Fall 1982): 339-357. [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] 101 NERA Economic Consulting [31] T. J. Holmes and J. A. Schmitz, "A theory of entrepreneurship and its application to the study of business transfers," Journal of Political Economy 89 (1990): 265-294. D. Holtz-Eakin, D. Joulfaian, and H. S. Rosen, "Entrepreneurial decisions and liquidity constraints," Journal of Political Economy 102 (1994): 53-75. D. Holtz-Eakin, D. Joulfaian, and H. S. Rosen, "Sticking it out: entrepreneurial survival and liquidity constraints," Rand Journal of Economics 25 no.2 (Summer 1994): 334-347. M. Hout and H. Rosen, "Self-Employment, family background, and race," Journal of Human Resources 35 no.4 (Fall 2000): 670-92. S. M. Ravi Kanbur, "Entrepreneurial risk taking, inequality, and public policy: an application of inequality decomposition analysis to the general equilibrium effects of progressive taxation," Journal of Political Economy 90 (1982): 1-21. M. Kidd, "Immigrant wage differentials and the role of self-employment in Australia," Australian Economic Papers 32 no.60 (June 1993): 92-115. R. E. Kihlstrom and J. J. Laffont, "A general equilibrium entrepreneurial theory of firm formation based on risk aversion," Journal of Political Economy 87 (1979): 719-848. P. J. Kuhn and H. J. Schuetze, "The dynamics of self-employment in Canada," working paper, McMaster University, 1998. A. Laferrere and P. McEntee, "Self-employment and intergenerational transfers of physical and human capital: An empirical analysis of French data," Economic and Social Review 27 no.1 (October 1995): 43-54. B. F. Lentz and D. N. Laband, "Entrepreneurial success and occupational inheritance among proprietors," Canadian Journal of Economics 23 (1990): 563-579. T. Lindh and H. Ohlsson, "Self-employment and windfall gains: evidence from the Swedish lottery," Economic Journal 106 no.439 (November 1996):1515-1526. J. E. Long, "The income tax and self-employment," National Tax Journal 35 (March 1982) 3142. N. Meager, "Does unemployment lead to self-employment?" Small Business Economics 4 (1992): 87-103. NERA Economic Consulting, "Minority Business Enterprise/Women Business Enterprise Availability Study," (2001). NERA Economic Consulting and Anchondo Research Management & Strategies, "Summary of MWBE Coverage Analysis," (2003). [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] 102 NERA Economic Consulting [46] A. R. Pickles and P. N. O'Farrell, "An analysis of entrepreneurial behavior from male work histories," Regional Studies 21 (1987): 425-444. J. F. Quinn, "Labor force participation patterns of older self-employed workers," Social Security Bulletin 43 (1980): 17-28. E. Reardon, "Are the self-employed misfits or superstars?" working paper, Rand Corporation (1998). H. Rees and A. Shah, "An empirical analysis of self-employment in the UK," Journal of Applied Econometrics 1 (1986): 95-108. M. T. Robson, "The rise in self-employment amongst UK males," Small Business Economics 10 no.3 (1998): 199-212 M. T. Robson, "Self-employment in the UK regions," Applied Economics 30 no.3 (1998): pp. 313-322. H. J. Schuetze, "Taxes, economic conditions and recent trends in male self-employment; a Canada-U.S. comparison," working paper, McMaster University, Hamilton, Ontario, Canada, 1998. M. P. Taylor, "Earnings, independence or unemployment; why become self-employed?" Oxford Bulletin of Economics and Statistics 58 no.2 (1996): 253-265. J. S. Wainwright, "Racial discrimination and minority business enterprise: evidence from the 1990 Census," Studies in Entrepreneurship Series. Edited by S. Bruchey. New York: Garland Publishing, 2000. [47] [48] [49] [50] [51] [52] [53] [54] 103