This is the accessible text file for GAO report number GAO-11-152 entitled 'Davis-Bacon Act: Methodological Changes Needed to Improve Wage Survey' which was released on April 6, 2011. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. United States Government Accountability Office: GAO: Report to the Chairman, Committee on Education and the Workforce, House of Representatives: March 2011: Davis-Bacon Act: Methodological Changes Needed to Improve Wage Survey: GAO-11-152: GAO Highlights: Highlights of GAO-11-152, a report to the Chairman, Committee on Education and the Workforce, House of Representatives. Why GAO Did This Study: Procedures for determining Davis-Bacon prevailing wage rates, which must be paid to workers on certain federally funded construction projects, and their vulnerability to the use of inaccurate data have long been an issue for Congress, employers, and workers. In this report, GAO examined (1) the extent to which the Department of Labor (Labor) has addressed concerns regarding the quality of the Davis- Bacon wage determination process, and (2) additional issues identified by stakeholders regarding the wage determination process. GAO interviewed Labor officials, representatives from contractor associations and unions, contractors, and researchers; conducted site visits to three Labor regional offices; and analyzed data from Labor’s wage survey database. What GAO Found: Recent efforts to improve the Davis-Bacon wage survey have not addressed key issues with timeliness, representativeness, and the utility of using the county as the basis for the wage calculation. Labor has made some data collection and processing changes; however, we found some surveys initiated under the new processes were behind Labor’s processing schedule. Labor did not consult survey design experts, and some criticisms of the survey and wage determination process have not been addressed, including the representativeness and sufficiency of the data collected. For example, Labor cannot determine whether its wage determinations accurately reflect prevailing wages because it does not currently calculate response rates or analyze survey nonrespondents. And, while Labor is required by law to issue wage rates by the “civil subdivision of the state,” the goal to issue them at the county level is often not met because of insufficient survey response. In the published results for the four surveys in our review, Labor issued about 11 percent of wage rates for key job classifications (types of workers needed for one or more of Labor’s construction types) using data from a single county. The rest were issued at the multi-county or state level. Over one-quarter of the wage rates were based on six or fewer workers. Figure: Percentage of Key Job Classification Wage Rates Issued at Each Geographic Level and Number of Employees Used to Determine Wage Rates, for Four Surveys Reviewed: [Refer to PDF for image: 2 pie-charts] Geographic level: State: 40%; Multi-county: 42%; County: 11%; Geographic level not available: 7%. Number of employees: 3 workers: 6%; 4-6 workers: 20%; 7-12 workers: 23%; 13-28 workers: 26%; 29 or more workers: 25%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in either 2009 or 2010. [End of figure] Little incentive to participate in Labor’s Davis-Bacon wage surveys and a lack of transparency in the survey process remain key issues for stakeholders. Stakeholders said contractors may not participate because they lack resources, may not understand the purpose of the survey, or may not see the point in responding because they believe the prevailing wages issued by Labor are inaccurate. While most stakeholders said the survey form was generally easy to understand, some identified challenges with completing specific sections. Our review of reports by Labor’s contracted auditor for four published surveys found most survey forms verified against payroll data had errors in areas such as number of employees and hourly and fringe benefit rates. Both contractor association and union officials said addressing a lack of transparency in how the published wage rates are set could result in a better understanding of the process and greater participation in the survey. What GAO Recommends: GAO suggests Congress consider amending its requirement that Labor issue wage rates by civil subdivision to allow more flexibility. To improve the quality and timeliness of the Davis-Bacon wage surveys, GAO recommends Labor obtain objective expert advice on its survey design and methodology. GAO also recommends Labor take steps to improve the transparency of its wage determinations. Labor agreed with the second recommendation, but said obtaining expert survey advice may be premature given ongoing changes. We believe obtaining expert advice is critical for improving the quality of wage determinations. View [hyperlink, http://www.gao.gov/products/GAO-11-152] or key components. For more information, contact Andrew Sherrill at (202) 512- 7215 or sherrilla@gao.gov. [End of section] Contents: Letter: Background: Recent Efforts to Improve Data Collection and Processing Have Not Yet Addressed Key Issues with Survey Quality: Little Incentive to Participate and Lack of Transparency Remain Key Issues for Stakeholders: Conclusions: Matter for Congressional Consideration: Recommendations for Executive Action: Agency Comments and Our Evaluation: Appendix I: Objectives, Scope, and Methodology: Appendix II: Wage Survey Form (WD-10): Appendix III: Labor's Wage Determination and Appeals Process under the Davis-Bacon Act: Appendix IV: Survey Announcement Letters Sent to Contractors and Interested Parties by the Department of Labor: Appendix V: Example of a Florida Wage Determination Published by the Department of Labor: Appendix VI: Comments from the Department of Labor: Appendix VII: GAO Contact and Staff Acknowledgments: Tables: Table 1: Percentage of Key Job Classification Wage Rates Issued at Each Geographic Level by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: Table 2: Percentage of Key Job Classification Wage Rates Issued Based on Number of Workers by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: Table 3: Percentage of Key Job Classification Wage Rates Using Federal Data by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: Figures: Figure 1: Flowchart of Labor's Wage Determination Process as of Fiscal Year 2010: Figure 2: Example of Labor Protocol for Combining Data from Miami-Dade County with Other Metropolitan Counties to Create Group, Supergroup, and State Wage Rates: Figure 3: Actual versus Expected Status of Eight Highway Surveys That Used Labor's New Processes, as of September 1, 2010: Figure 4: Actual versus Expected Status of Four Building and Heavy Surveys That Used Labor's New Processes, as of September 1, 2010: Figure 5: Percentage of Key Job Classification Wage Rates Issued at Each Geographic Level, for Four Surveys Reviewed: Figure 6: Percentage of Key Job Classification Wage Rates Issued Based on Number of Workers, for Four Surveys Reviewed: Figure 7: Labor's Printed WD-22 Report for a Texas Building Survey: Figure 8: Excerpt from a Florida Wage Determination Published on Labor's Web Site: Abbreviations: ABC: Associated Builders and Contractors, Inc. AGC: Associated General Contractors of America: ASDS: Automated Survey Data System: BLS: Bureau of Labor Statistics: CBA: collective bargaining agreement: CIRPC: Construction Industry Research and Policy Center: IBEW: International Brotherhood of Electrical Workers: Labor: Department of Labor: MSA: metropolitan statistical area: OIG: Office of Inspector General: OMB: Office of Management and Budget: Recovery Act: American Recovery and Reinvestment Act of 2009: WDGS: Wage Determination Generation System: [End of section] United States Government Accountability Office: Washington, DC 20548: March 22, 2011: The Honorable John Kline: Chairman: Committee on Education and the Workforce: House of Representatives: Dear Mr. Chairman: Procedures for determining Davis-Bacon prevailing wage rates, which must be paid to workers on certain federally funded construction projects, have long been an issue for Congress, employers, and workers. Concerns have focused on the Department of Labor's (Labor) procedures for determining prevailing wages and their vulnerability to the use of inaccurate data. In the 1990s, we issued two reports that found process changes were needed to increase confidence that wage rates were based on accurate data.[Footnote 1] A third report, issued in 1999, found changes planned by Labor, if successfully implemented, had the potential to improve the survey process used to determine local prevailing wages.[Footnote 2] However, in 2004, Labor's Office of Inspector General (OIG) found Labor's approaches had not resolved past concerns, and wage data errors and the timeliness of surveys continued to be issues.[Footnote 3] More recently, the passage of the American Recovery and Reinvestment Act of 2009 (Recovery Act)[Footnote 4] focused attention on the need for accurate and timely wage determinations, with more than $300 billion estimated to come from Division A of the act which provides substantial funding for, among other things, federally funded building and infrastructure work potentially subject to Davis-Bacon wage rates.[Footnote 5] To address these issues, you asked us to examine Labor's current implementation of the Davis-Bacon Act. We assessed (1) the extent to which Labor has addressed concerns regarding the quality of the Davis- Bacon wage determination process and (2) additional issues identified by stakeholders regarding the wage determination process. To assess the extent to which Labor has addressed concerns regarding the quality of the wage determination process, we reviewed relevant federal laws and regulations, interviewed Labor officials, and reviewed agency documents on current survey practices and compared them with guidance on data quality and survey design from the Office of Management and Budget (OMB) and Labor. OMB designed its guidance on data quality to ensure agencies meet basic information quality standards for objectivity, utility, and integrity of information before it is disseminated.[Footnote 6] We conducted site visits to three of Labor's five regional offices that process Davis-Bacon wage surveys--Northeast region (Philadelphia), Southeast region (Atlanta), and Southwest region (Dallas)--to interview regional staff who, Labor officials said, were conducting surveys under Labor's recently revised processes.[Footnote 7] In addition, we conducted a site visit to the Construction Industry Research and Policy Center (CIRPC) at the University of Tennessee, which is contracted by Labor to estimate the potential number of construction projects to be included in a specific survey and conduct some aspects of the survey process. We also compared 12 surveys for specific types of construction projects performed under Labor's new processes against its revised processing timelines to assess whether the surveys were on schedule. Further, we analyzed data from Labor's Automated Survey Data System (ASDS) for wage rates from the 2005 surveys of Florida and Maryland and the 2006 surveys of Tennessee and West Texas Metropolitan. We selected these surveys because results were recently published (in 2009 or 2010) and they represented geographic diversity to the extent possible.[Footnote 8] We conducted analyses on the geographic level at which rates were issued (i.e., county, group, supergroup, or statewide) and the number of workers used to calculate wage rates. In addition, we analyzed data from Labor's Wage Determination Generation System to determine the proportion of union-prevailing to nonunion-prevailing wage rates and the age of currently published wage rates. We assessed the reliability of the data we used by reviewing pertinent system and process documentation, interviewing knowledgeable officials, and conducting electronic testing on data fields necessary for our analysis. We found the data we reviewed reliable for the purposes of our analysis. To assess what additional issues were concerns for stakeholders, we conducted approximately 30 interviews with a nonprobability selection of representatives from academia, contractor associations, contractors, and unions and performed a content analysis of their responses. For more information on our scope and methodology, see appendix I. We conducted this performance audit from September 2009 through March 2011 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Background: The Davis-Bacon Act[Footnote 9] was enacted in 1931, in part, to protect communities and workers from the economic disruption caused by contractors hiring lower-wage workers from outside their local area, thus obtaining federal construction contracts by underbidding competitors who pay local wage rates. Davis-Bacon generally requires employers to pay locally prevailing wages and fringe benefits to laborers and mechanics employed on federally funded construction projects in excess of $2,000.[Footnote 10] The Recovery Act requires all laborers and mechanics employed by contractors and subcontractors on projects funded directly or assisted by the federal government through the Recovery Act also be paid at least the prevailing wage rate under Davis-Bacon.[Footnote 11] Our previous work found 40 programs, such as the Weatherization Assistance Program, newly subject to Davis-Bacon requirements as a result of the Recovery Act's prevailing wage provision.[Footnote 12] Of these, 33 programs existed prior to the Recovery Act but were subject to the Davis-Bacon requirements for the first time, and 7 were newly created programs. In 2009, federally funded construction and rehabilitation, including projects funded through the Recovery Act, totaled about $220 billion. [Footnote 13] Labor administers the Davis-Bacon Act through its Wage and Hour Division, which conducts voluntary surveys of construction contractors and interested third parties[Footnote 14] on both federal and nonfederal projects to obtain information on wages paid to workers in each construction job classification by locality. It then uses the data submitted on these survey forms to determine local prevailing wage and fringe benefit rates. In 2002, Labor began conducting simultaneous statewide surveys for all four of its construction types: highway, residential, building, and heavy.[Footnote 15] Labor describes highway construction as the construction, alteration, or repair of roads, streets, highways, runways, alleys, trails, parking areas, and other similar projects not incidental to building or heavy construction. Residential construction includes single-family homes and apartment buildings that are not more than four stories. If a structure that houses people is over four stories or if it houses machinery, equipment, or supplies, it is considered building construction. Heavy construction generally includes any project that does not fall into the other three categories--for example, dam and sewer projects. Labor determines which states it will survey each year based on a variety of factors, including the date of a state's most recent survey, planned federal construction, and complaints or requests from interested parties on current wage determinations. The calculated wage and fringe benefit rates that result from the surveys are posted online in wage determinations and used by contractors working on federal construction projects to prepare bids and pay workers. Past Concerns and Efforts to Improve the Wage Determination Process: Both GAO and the Labor OIG have reported concerns with Labor's wage determination process. In 1996, we found Labor had internal control weaknesses that contributed to lack of confidence in the wage determinations, including limitations in Labor's verification of wage and fringe benefit data, its computer capabilities, and an appeals process that was difficult for interested parties to access.[Footnote 16] In 1997, the OIG found much of the data it examined to be inaccurate and potentially biased due to weaknesses in survey methodology.[Footnote 17] For fiscal year 1997, Congress directed $3.75 million toward improvements to the wage determination process. Using five criteria-- feasibility/viability, timeliness, accuracy, completeness, and cost-- Labor evaluated two options: * Reengineering: Apply new technologies and processes to the existing Davis-Bacon survey program to increase participation in and improve the accuracy and timeliness of the surveys. * Reinvention: Use existing Bureau of Labor Statistics (BLS) data, specifically data from BLS's Occupational Employment Statistics survey and National Compensation Survey, as the primary basis for Davis-Bacon wage determinations.[Footnote 18] In 1999, as Labor was evaluating these options, we again reviewed the wage determination process and found, in response to a directive from a congressional committee and our recommendation, Labor had implemented a program to verify a sample of wage survey data, including verifying data on site using employer payrolls.[Footnote 19] However, we agreed with the OIG that verification efforts be viewed as temporary steps until more fundamental reforms could be made to Labor's survey methodology. We also found that reengineering or reinvention had the potential to improve the accuracy and timeliness of the wage determination process.[Footnote 20] In January 2001, Labor reported to Congress it would pursue reengineering. Labor concluded that reinvention (using BLS data) would have the benefits of accuracy and timeliness, but presented challenges, including difficulty in determining fringe benefits and in producing wage estimates for a broad range of construction job classifications. Reengineering, which included improvements to the wage survey form (including a scannable form and online version) and a computer system to assist with data clarification and analysis, would make it feasible to survey every area of the country for all four construction types no less than every 3 years, Labor concluded. In 2004, the Labor OIG found Labor's reengineering had not resolved past concerns. In a sample of wage survey forms (known as WD-10s) from before and after reengineering, the OIG found errors in almost 100 percent of verified survey forms. The OIG said these errors occurred even with a revised WD-10, the introduction of an online WD-10, and efforts by Labor analysts to review and correct data. Mistakes in survey data included respondents using incorrect peak weeks,[Footnote 21] miscounts in the number of workers in each job classification, and misreporting of wage rates--for example, reporting one wage rate for a job classification when two or more wage rates existed. In addition, the OIG reported concerns about bias because only contractors with the personnel to complete WD-10s may respond and some may not participate to avoid involvement with the government. The OIG also found that higher participation by either unions or nonunion contractors could potentially weight the wage and benefit rates in their favor. Finally, the OIG noted there had been little improvement since its 1997 review in the time required to issue wage determinations. Current Survey Process: The current survey process, which conducts statewide surveys for all construction types, consists of five basic phases (see figure 1). [Footnote 22] Figure 1: Flowchart of Labor's Wage Determination Process as of Fiscal Year 2010: [Refer to PDF for image: illustration] Stage 1: Planning; * Plan survey activity; * Establish survey criteria; * Obtain list of construction projects. Stage 2: Survey; * Announce survey; * Conduct pre-survey briefings; * Collect data. Stage 3: Data clarification and analysis; * Clarify questionable information received; * Calculate wage and fringe benefit rates; * Determine sufficiency; * Verify a sample of data; * Update data based on verification and calculate final rates; * Transmit results from regional office to Labor headquarters. Stage 4: Issuance; * Review results/recommend action; * Issue wage determination. Stage 5: Appeals; * Reconsideration and review ongoing after issuance. Source: GAO analysis of Labor’s wage determination process. [End of figure] Prior to the start of a survey, Labor identifies the state, construction types, and survey time frame--the time period in which a construction project needs to be active to meet survey criteria--and requests that CIRPC provide a report on active construction projects for the identified time frame, construction type, and geographical area. F.W. Dodge Reports for those projects are then ordered and reviewed to ensure they meet the basic criteria of the survey. [Footnote 23] Once a survey is scheduled, Labor usually conducts pre- survey briefings for interested parties to clarify survey procedures and provide information on how data should be submitted. Labor then sends surveys to general contractors identified through the Dodge Reports and relevant interested parties in the area to be surveyed. (See appendix II for a copy of the wage survey.) It also requests information from federal agencies on construction projects that meet survey criteria. A follow-up letter is sent to general contractors who do not respond. Subcontractors, identified by the general contractors, are also sent an initial letter with a survey and a follow-up letter if they do not respond. Completed wage survey forms are returned by either contractors or interested parties and are reviewed, under a contract with Labor, by CIRPC, which matches submitted information with its construction project and forwards it to the appropriate Labor regional office.[Footnote 24] The regional offices clarify missing, ambiguous, or inconsistent information to the extent possible, and pull random samples of wage survey forms to verify by phone or on site. Officials request that supporting payroll documentation be sent to the regional office. For on-site verification, Labor contracts with a private accounting firm whose auditors review payroll records. Any discrepancies between the wage survey form and the contractor's payroll records are reviewed and corrected in the survey data by Labor regional staff. Contractors selected for verification, who are not able or willing to provide payroll records, can still be included in the survey in most cases.[Footnote 25] See appendix III for a more detailed description of the wage determination process. Labor uses several procedures to calculate wage rates and determine if it has sufficient information from collected and verified surveys to issue a wage determination--a compilation of prevailing wage rates for multiple job classifications in a given area. In determining a prevailing wage for a specific job classification, Labor considers sufficient data to be the receipt of data on at least three workers from two different employers in its designated area who have that job. Then, in accordance with its regulations, Labor uses a "50-percent rule" to calculate the prevailing wage. The 50-percent rule states the prevailing wage is the wage paid to the majority (over 50 percent) of workers employed in a specific job classification on similar projects in the area.[Footnote 26] If the same rate is not paid to a majority (over 50 percent) of workers in a job classification, the prevailing wage is the average wage rate weighted by the number of employees for which that rate was reported. In cases where the prevailing rate is also a collectively bargained, or union, rate, the rate is determined to be "union-prevailing." According to Labor's policy, union- prevailing wage rates in wage determinations can be updated when there is a new collective bargaining agreement (CBA) without Labor conducting a new survey. Nonunion-prevailing wage rates are not updated until a new survey is conducted. To issue a wage determination for a construction type in a given area, Labor must, according to its procedures, also have sufficient data to determine prevailing wages for at least 50 percent of key job classifications. Key job classifications are those determined necessary for one or more of the four construction types.[Footnote 27] By statute, Labor must issue wage determinations based on similar projects in the "civil subdivision of the state" in which the federal work is to be performed.[Footnote 28] Labor's regulations state the civil subdivision will be the county, unless there are insufficient wage data.[Footnote 29] When data from a county are insufficient to issue a wage rate for a job classification, a group of counties is created by combining a rural county's data with data from one or more contiguous rural counties. A metropolitan county's data are combined with data from other counties in the state within the metropolitan statistical area (MSA). If data are still insufficient to issue a wage rate, a supergroup is created by combining a rural county's data with data from additional contiguous rural counties, or a metropolitan county's data are combined with county data from other MSAs or the consolidated MSA counties. Finally, if this supergroup still does not provide sufficient wage data to issue a wage rate for a job classification, a statewide rate is created by combining data for all rural counties or all metropolitan counties in the state. Counties are combined based on whether they are metropolitan or rural, and cannot be mixed.[Footnote 30] Once wage determinations are issued, an interested party may seek reconsideration and review through an appeals process.[Footnote 31] See figure 2 for an example of how wage data from Miami-Dade County, Florida, are combined, as needed, with data from other counties to create group, supergroup, and state wage rates. Figure 2: Example of Labor Protocol for Combining Data from Miami-Dade County with Other Metropolitan Counties to Create Group, Supergroup, and State Wage Rates: [Refer to PDF for image: illustrated map of Florida] County (1 county): Miami-Dade. Group (3 counties): Broward; Miami-Dade; Palm Beach. Supergroup (5 counties): Broward; Martin; Miami-Dade; Palm Beach; St. Lucie. State (37 counties): Alachua; Baker; Bay; Brevard; Broward; Charlotte; Clay; Collier; Duval; Escambia; Gadsden; Gilchrist; Hernando; Hillsborough; Indian River; Jefferson; Lake; Lee; Leon; Manatee; Marion; Martin; Miami-Dade; Nassau; Orange; Osceola; Palm Beach; Pascoe; Pinellas; Polk; Santa Rosa; Sarasota; Seminole; St. John's; St. Lucie; Volusia; Wakulla. Source: GAO analysis of Labor data; Map Resources (map). [End of figure] Recent Efforts to Improve Data Collection and Processing Have Not Yet Addressed Key Issues with Survey Quality: Labor's Changes to Survey Data Collection and Processing May Not Achieve Expected Results: Labor has taken several steps over the last few years to address issues with its Davis-Bacon wage surveys, including completing a number of open surveys and changing how it collects and processes some survey data in its efforts to improve timeliness and accuracy. However, these efforts may not achieve Labor's desired results. We found some surveys initiated under the new process are behind schedule and some published wage rates are based on outdated data. In 2007, Labor officials decided not to initiate any new surveys in order to finalize and publish results from 22 open surveys, which accumulated after Labor began conducting statewide surveys in 2002. [Footnote 32] Regional office officials said it was difficult and time- consuming to clarify and verify data in these surveys because contractors often did not have easy access to records for survey data which, in some cases, had been submitted several years earlier. As of September 1, 2010, results from 20 of the 22 surveys were published and results from the remaining 2 were in the process of being published. Officials said once results from all 22 surveys are published, they will be able to focus on more recent surveys, which will reduce delays in processing and increase accuracy because more recently collected information is easier and less time-consuming to clarify and verify with contractors. Labor also changed how it collects survey data for its four construction types after it conducted an informal review in 2009. Labor officials said they had been using a "one size fits all" approach to surveys and were not accounting for differences in types of construction activity, the demographic characteristics of a given state, and available sources of wage data. To address these differences, Labor began surveying some of its four construction types separately instead of surveying all construction types simultaneously in a given state. Labor also began using certified payrolls as the primary data source for highway surveys. Labor officials said most highway construction has a federal component and certified payrolls provide accurate and reliable wage data.[Footnote 33] Officials also said using certified payrolls eliminates the need for on-site verification of reported wage data, although Labor continues to survey interested parties.[Footnote 34] Officials estimate these efforts will reduce processing time for highway surveys by more than 80 percent, or from about 42 months to 8 months. Labor adjusted its survey processes for residential, building, and heavy construction types as well. For surveys of residential construction, Labor plans to phone contractors and unions and visit contractor associations to increase a historically low response. Officials said these collection methods will be possible because of the small number of residential projects compared to other construction types. Labor began conducting a new residential survey in 2010. For building and heavy construction, Labor started a pilot with five surveys in 2009,[Footnote 35] adjusting survey time frames--the time period in which a construction project has to be active for it to meet survey criteria--to better manage the quantity of data received. Labor found its previous 1-year survey time frame produced, in some cases, too many or too few responses for building and heavy surveys. Instead, by adjusting the survey time frame to account for the number of projects in a particular region (with shorter time frames for areas in which there are many active projects), Labor expects to reduce the time needed to process surveys and determine prevailing wages. Overall, Labor estimates these changes will reduce processing time for building and heavy surveys by approximately 54 percent, or from about 37 months to 17 months. Labor also revised its approach to processing data for all surveys. Labor's regional offices began reviewing and analyzing survey forms when they are received rather than waiting until a survey closes. Labor officials said this processing of data in "real time" will improve timeliness and accuracy because survey respondents will be better able to recall the submitted information when contacted by regional office staff for clarification and verification. While it is too early to fully assess the effects of Labor's 2009 changes, our review found timeliness is still an issue and improvements expected from processing changes may not be fully realized. Of the 16 surveys started under Labor's new processes at the time of our review, we were unable to analyze the timeliness of 4--3 highway surveys and 1 building and heavy survey--because of unclear dates in Labor's data. A senior Labor official said regional offices differed as to when they recorded dates for key survey activities, and we found some recorded dates were out of sequence. During the course of our review, the senior Labor official said regional offices will consistently enter key dates for future surveys, which will allow Labor to better assess whether new processes are improving timeliness. Of the remaining 12 surveys for which we were able to assess timeliness, 8 were highway surveys for which Labor requested certified payrolls. Of those 8, we found 6 were behind schedule, 1 was on schedule, and 1 had not started as of September 1, 2010 (see figure 3).[Footnote 36] A senior Labor official said staff did not immediately start processing all certified payrolls--requested for all federal projects within a specific 1-year period--when they were received because of regional office workloads. As a result, some certified payroll data were months old before Labor surveyed interested parties. For example, as of September 1, 2010, certified payroll data for the Florida 2009 highway survey were 8 months old, though Labor had not yet surveyed interested parties. Moreover, processing certified payrolls may be labor-intensive and time- consuming. A senior Labor official said the agency cannot predict how many certified payrolls will be submitted by state departments of transportation and often receives boxes of documents for each survey. Some regional office officials said extracting information from certified payrolls is difficult because of inconsistent formats and frequently requires clarification with contractors. To address these potential delays, a senior Labor official said they are considering collecting certified payrolls monthly from states with upcoming surveys, and processing the payrolls as they are received. Figure 3: Actual versus Expected Status of Eight Highway Surveys That Used Labor's New Processes, as of September 1, 2010: [Refer to PDF for image: illustrated table] Highway survey: Florida 2009[A]; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): [Empty]. Highway survey: New Mexico 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: Actual survey activity, as of September 1, 2010; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Highway survey: North Carolina 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: Actual survey activity, as of September 1, 2010; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Highway survey: Oklahoma 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: Actual survey activity, as of September 1, 2010; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Highway survey: South Carolina 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: Actual survey activity, as of September 1, 2010; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Highway survey: Louisiana 2010; Survey activity: Pre-survey activity: [Empty]; Survey preparation: Actual survey activity, as of September 1, 2010; Data collection (from certified payrolls and interested party survey forms): Expected survey activity given number of days in progress; Data clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): [Empty];. Highway survey: Nebraska 2010; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: Actual survey activity, as of September 1, 2010; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Highway survey: New Hampshire 2010; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection (from certified payrolls and interested party survey forms): [Empty]; Data clarification and analysis: Expected survey activity given number of days in progress; Actual survey activity, as of September 1, 2010; Contractor and interested party data verification: [Empty]; On-site verification of interested party data (if necessary): [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): [Empty]. Source: GAO analysis of Labor data. Note: Expected survey activity was determined using the maximum number of days allotted for each activity in Labor's new timeline. Actual survey activity was based on Labor's estimated date that regional offices entered the survey in ASDS. [A] The Florida 2009 highway survey had not started as of September 1, 2010. [End of figure] The remaining 4 surveys were building and heavy surveys and all were behind schedule as of September 1, 2010 (see figure 4). Figure 4: Actual versus Expected Status of Four Building and Heavy Surveys That Used Labor's New Processes, as of September 1, 2010: [Refer to PDF for image: illustrated table] Building and heavy survey: Montana 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection, clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site data verification: Actual survey activity, as of September 1, 2010; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Building and heavy survey: North Carolina 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection, clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site data verification: Actual survey activity, as of September 1, 2010; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Building and heavy survey: West Virginia 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection, clarification and analysis: Actual survey activity, as of September 1, 2010; Contractor and interested party data verification: [Empty]; On-site data verification: [Empty]; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Building and heavy survey: Wyoming 2009; Survey activity: Pre-survey activity: [Empty]; Survey preparation: [Empty]; Data collection, clarification and analysis: [Empty]; Contractor and interested party data verification: [Empty]; On-site data verification: Actual survey activity, as of September 1, 2010; Area practice adjustment, prevailing wage calculation, and rate recommendation: [Empty]; National office review and publication of rates: [Empty]; Post-publication (includes reconsideration and review, if necessary): Expected survey activity given number of days in progress. Source: GAO analysis of Labor data. Note: Expected survey activity was determined using the maximum number of days allotted for each activity in Labor's new timeline. Actual survey activity was based on Labor's estimated date that regional offices entered the survey in ASDS. [End of figure] In conducting a "universe" or "census" survey of all active construction projects within a designated time frame and area, Labor accepts data from a variety of sources, including contractors and interested parties. As a result, the number of returned survey forms and the time required to clarify data can vary widely. For example, for 14 surveys conducted under past processes, the number of survey forms received for each ranged from less than 2,000 to over 8,000, and the average processing time for data clarification and analysis ranged from 10 months to more than 40.[Footnote 37] After the 2009 changes, Labor estimates survey data clarification and analysis will take about 1 to 7 months, depending on construction type.[Footnote 38] Some of the anticipated time savings, particularly for building and heavy surveys, is based on managing fewer forms because of its focus on the number of projects in a particular region rather than a 1-year time frame. However, by accepting data submitted by contractors and interested parties on any relevant project as part of its universal survey approach, Labor is limited in its ability to predict how many forms will be returned and the time needed to process them. The more time required, the more likely wage rates will be outdated when published in wage determinations. In addition, Labor cannot entirely control when it receives survey forms. Though Labor officials said processing survey forms as they are received will improve timeliness, some regional office officials told us this "real time" processing approach has a limited effect because the bulk of the forms are returned on the last day of a survey. Additionally, officials in two of the three regional offices we visited said this new approach is not substantially different from their previous procedure. Since our site visits, a senior Labor official said analysts at regional offices have noticed a difference between processing forms in "real time" and their previous procedure, and that increased use of online submissions is expected to help reduce last-minute survey returns. To address such challenges, OMB guidance suggests agencies consider the benefits and costs of conducting a sample survey instead of a census survey. [Footnote 39] According to OMB, a sample can be used to ensure data quality in a way that is often more efficient and economical than a census. The fact that Labor is behind schedule on surveys even with the new 2009 processes may affect the agency's ability to update the many published nonunion-prevailing wage rates, which are several years old. Labor's fiscal year 2010 performance goal was for 90 percent of published wage rates for building, heavy, and highway construction types to be no more than 3 years old. Our analysis of published rates for these three construction types found 61 percent were 3 years old or less as of November 12, 2010. However, this figure is somewhat misleading because it includes both union-prevailing and nonunion- prevailing wage rates, which differ in how they are updated. Union- prevailing rates, which constitute almost two-thirds of the over 650,000 published building, heavy, and highway rates, may be updated when new CBAs are negotiated, and we found almost 75 percent of those rates were 3 years old or less as of November 12, 2010. However, 36 percent of nonunion-prevailing rates, which are not updated until Labor conducts a new survey, were 3 years old or less, and almost 46 percent were 10 or more years old. One regional office official and two stakeholders we interviewed said Labor, in some cases, has had to update nonunion-prevailing rates without a new survey because they no longer complied with the federal minimum wage. Moreover, wage rates at the time of publication may reflect wage data from several years prior due to processing delays. For example, of the 20 open surveys for which Labor had published results as of September 1, 2010, 9 published in 2009 or 2010 were based on data 5 or more years old at the time of publication and, of those, 3 were based on data 7 or more years old. [Footnote 40] Though these survey results were only recently published, the age of the wage data they contain means those states will likely need to be resurveyed soon. Several of the union and contractor association officials we interviewed said the age of the Davis-Bacon nonunion-prevailing rates means they often do not reflect actual prevailing wages. As a result, they said it is more difficult for both union and nonunion contractors to successfully bid on federal projects because they cannot recruit workers with artificially low wages but risk losing contracts if their bids reflect more realistic wages. Labor officials said the only way to correct the age disparity between union-and nonunion-prevailing rates is to conduct surveys more frequently; however, some regional office officials said the goal to survey each area every 3 years is not feasible with current processes. Those who said it is feasible cited the need for adequate technology and staffing, which they said is not in place in all regional offices. Critical Problems with Labor's Wage Survey Methodology Still Hinder Quality: Although Labor has made recent changes to data collection and processing, some critical problems with its survey methodology have not been addressed. Our review identified persisting shortcomings in the representativeness of survey results and the sufficiency of data gathered for Labor's county-focused wage determinations. Representativeness: OMB guidance states that agencies need to consider the potential impact of response rate and nonresponse on the quality of information obtained from a survey, and suggests agencies consult with trained survey methodologists when designing surveys to address this issue. Rather than conducting a formal evaluation of the wage survey process and consulting with experts in survey design and methodology, a senior Labor official said the agency based changes on an informal review that drew on staff experiences. While our prior work has shown it is reasonable and desirable to obtain input from knowledgeable staff, technical guidance from experts is considered critical to ensure the validity and reliability of survey results. Labor cannot determine whether its Davis-Bacon survey results are representative of prevailing wage rates because it does not currently calculate response rates or conduct a nonresponse analysis. According to OMB, response rate calculation and nonresponse analysis are important because a low response rate may mean survey results are misleading or inaccurate if those who respond to a survey differ substantially and systematically from those who do not respond. A Labor official said that when the agency started conducting statewide surveys in 2002, it stopped calculating overall response rates because of the large volume of data received and challenges in tracking who submitted specific information. In addition, the official said Labor could not collect enough data to meet its then-standard of data on at least six workers from three different employers for each job classification, so it changed the standard to its current three workers from two employers.[Footnote 41] This standard can be met using data from a single county, multiple counties within a state, or statewide. Also, aside from a second letter sent automatically to survey nonrespondents, Labor does not currently have a program to systematically follow up with or analyze all nonrespondents. Labor's own procedures manual recognizes nonresponse as a potential source of survey bias and indicates there is a higher risk nonrespondents will be nonunion contractors because they may have greater difficulty in compiling wage information or be more cautious about reporting wage data. Despite this guidance, regional office officials said they spend the bulk of their time clarifying data received. Of Labor's published wage rates as of November 12, 2010, about 63 percent were union- prevailing; in contrast, about 14 percent of construction workers nationwide were represented by unions in 2010, according to BLS figures[Footnote 42]. Several of the stakeholders we interviewed said the fact that Labor does not ensure the representativeness of the survey responses reduces the accuracy of the published wage rates. In addition, some regional office officials said statistical sampling may make wage rates more accurate, although they cautioned that some contractors or interested parties may not support a change to sampling if it meant they would be excluded from participating in the survey. During the course of our review, a senior official said Labor is taking steps to again calculate response rates, beginning with updates to the survey database and changes to the survey form, which will more clearly identify who submitted wage information.[Footnote 43] However, because Labor has not yet fully implemented these changes, it is unclear if they will lead to improving the quality of the survey. Utility of County Focus: Although its regulations state the county will normally be the civil subdivision for which a prevailing wage is determined,[Footnote 44] Labor is often unable to issue wage rates for job classifications at the county level because it does not collect enough data to meet its current sufficiency standard of wage information on at least three workers from two employers. In the results from the four surveys we reviewed--Florida 2005, Maryland 2005, Tennessee 2006, and West Texas Metropolitan 2006--Labor issued about 11 percent of wage rates for key job classifications using data from a single county (see figure 5). About 22 percent of the wage rates were issued at the group level (combined data from a group of counties within the same state) and about 20 percent at the supergroup level (combined data from other groups of counties within the same state). Almost 40 percent of the wage rates were issued at the statewide level incorporating data from either all metropolitan or all rural counties in the state.[Footnote 45] The remaining 7 percent were issued for combined counties for which the geographic calculation level was not available.[Footnote 46] (For more information on how the geographic level of issued wage rates varied by construction type and by metropolitan and rural rates, see appendix I.) Figure 5: Percentage of Key Job Classification Wage Rates Issued at Each Geographic Level, for Four Surveys Reviewed: [Refer to PDF for image: pie-chart] Geographic level: State: 40%; Multi-county: 42%; County: 11%; Geographic level not available: 7%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in either 2009 or 2010. [End of figure] In 1997, Labor's OIG reported that issuing rates by county may cause wage decisions to be based on an inadequate number of responses. In our review of the four surveys, we found one-quarter of the final wage rates for key job classifications were based on wages reported for six or fewer workers (see figure 6). (For more information on how the number of workers used to determine rates varied by construction type and by metropolitan and rural rates, see appendix I.) Figure 6: Percentage of Key Job Classification Wage Rates Issued Based on Number of Workers, for Four Surveys Reviewed: [Refer to PDF for image: pie-chart] Number of employees: 3 workers: 6%; 4-6 workers: 20%; 7-12 workers: 23%; 13-28 workers: 26%; 29 or more workers: 25%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in either 2009 or 2010. [End of figure] In the surveys we reviewed, we also found Labor sometimes determined prevailing wages based on small amounts of data even in metropolitan areas. For example, in the 2005 survey of building construction in Florida, the prevailing wage rate for a forklift operator in Miami- Dade County was based on wages reported for five workers statewide. The statutory requirement to issue Davis-Bacon prevailing wages based on a "civil subdivision of the state"[Footnote 47] also limits Labor's options to address inadequate data. For example, Labor is not able to augment its survey data with data from other sources because those sources may draw from other geographic areas, such as MSAs, which are not the same as civil subdivisions.[Footnote 48] Officials from Labor's survey contractor, CIRPC, said one way to improve accuracy is to survey areas other than counties. CIRPC officials said the current wage survey uses arbitrary geographic divisions, in contrast to other groupings, such as the economic areas used by the Bureau of Economic Analysis, which are based on relevant regional markets that frequently cross county and state lines.[Footnote 49] These groupings, they said, are more reflective of area wage rates. Some stakeholders said the focus on county-level wage rates results in the publication of illogical rates. One contractor association representative said metropolitan statistical areas would be more appropriate in New York, for example, because there is a larger difference in wages between upstate and downstate New York than between the counties containing the cities of Rochester, Syracuse, and Buffalo. Another contractor association representative said the geographic divisions used by Labor for prevailing wages are illogical for projects not confined to a single county, offering the example of a contractor paving a road that crossed a county line and who was forced to pay workers different wage rates based on which side of the line they worked. Little Incentive to Participate and Lack of Transparency Remain Key Issues for Stakeholders: Many Stakeholders Reported Contractors Lack Incentive to Participate in Davis-Bacon Surveys: In our interviews with stakeholders about additional issues with Labor's wage determination process, they provided several reasons why contractors have little or no incentive to participate in the Davis- Bacon wage survey. First, 19 of 29 stakeholders said contractors may not have the time or resources to respond.[Footnote 50] An employee for one contractor said she had returned the wage survey but might not have had she known it was voluntary because her company was short- staffed. Other stakeholders said contractors might not see the survey as a priority. Second, 16 stakeholders said contractors either may not understand the purpose of the survey or do not see the point in responding because they believe the prevailing wages issued by Labor are inaccurate. Third, 10 stakeholders said contractors may be reluctant to provide information to the government because they view it as proprietary or fear that doing so will subject them to audits. Finally, 8 stakeholders said contractors who do not work on public projects may not understand the survey is soliciting wage data from private as well as public projects so they do not think they need to respond. For instance, representatives from one state contractor association said some contractors believe the wage survey only serves to perpetuate established rates because wage surveys sent by Labor may have the names of projects subject to Davis-Bacon already entered on the form.[Footnote 51] Officials we interviewed in Labor regional offices echoed many of these concerns. They said contractors either think their survey responses will not make a difference in the determination of prevailing wages or are unaware they are being asked to submit information on private projects. A contributing factor, one official said, is that the survey announcement letter may not clearly communicate it is soliciting information on both public and private construction. In our review of the contractor announcement letter, we found it states that requested information will be used to set prevailing wages and asks the contractor to fill out the wage survey for the construction project listed on the form and any additional projects that fit survey criteria. But the letter does not specifically state that Labor is soliciting data for both public and private projects. (See appendix IV for copies of the survey announcement letters sent to contractors and interested parties.) Additionally, some regional office officials said larger contractors may be more likely to respond because they have more resources, including administrative personnel, to complete the survey form. They said contractors also may not respond because they find the form complicated or do not understand its importance. Yet if contractors call the regional office and Labor staff have an opportunity to explain the reason for the survey and answer questions, many of those callers seem more receptive to participating, some regional office officials said. A lack of survey participation by those on private construction projects could result in Labor having to use data from federal projects, which are already paying Davis-Bacon wages, to set prevailing wages for building and residential construction. Per its regulations, Labor uses federal project data in all highway and heavy surveys, but it only uses federal project data in building and residential surveys when it lacks sufficient data from nonfederal projects.[Footnote 52] In the results from the four surveys we reviewed, almost one-quarter of the building wage rates and over two- thirds of the residential rates for the 16 key job classifications, such as carpenter and common laborer, included federal data. (For more information on how the percentage of federal data varied by metropolitan and rural rates, see appendix I.) Stakeholders Reported Survey Form Was Generally Easy to Understand, but Most Forms Reviewed Had Errors: While 19 of the 27 contractors and interested parties we interviewed said the wage survey form, which Labor officials said was last updated in 2004, is generally easy to understand, some identified challenges in completing specific sections.[Footnote 53] For example, five stakeholders said it is difficult to know which job classification applies to their workers. Representatives from one national contractor association said they had previously informed Labor the survey form does not reflect nonunion industry practices and contractors may not track data in a way that makes it easy to fill out the form. As a result, they said most nonunion contractors opt not to return the wage survey rather than attempt to break down their data to fit its format. Other state contractor association representatives said workers on some construction sites today perform tasks across multiple job classifications; for example, a carpenter may also perform some tasks of a laborer. Yet the survey form asks contractors to provide wages for a worker by a single job classification. In addition, officials from one state local union said, to assist contractor participation in the survey, they created and distributed their own spreadsheet for contractors to fill out because they thought it would be more easily understood than Labor's wage survey form. Labor reported to Congress in 2006 that use of the scannable survey form resulted in submission of more complete data, but our analysis of reports for four state surveys found most verified forms still had errors.[Footnote 54] During on-site verification, Labor's contracted accounting firm compares clarified wage survey data to a sample of contractor payroll records and reports any discrepancies. These auditor reports show mistakes occurred most often in the number of employees reported in each job classification, listed hourly and fringe benefit wage rates, and project dollar value, some of which were also issues in the 2004 Labor OIG report. A senior Labor official said one reason contractors make errors on the form may be because they fill it out from memory rather than consulting their payroll records. Officials said they expect such errors to decrease under the new survey processes as Labor analysts clarify contractor-submitted data sooner. Some of these errors may be due to the fact that Labor did not pretest its current survey form with respondents. Officials said they are planning another update to address portions of the form that consistently confuse respondents. These include not having a place to note an "interested party," rather than a "contractor" or "subcontractor," is filling out the survey form, as well as improvements to the section on job classifications and fringe benefits. Labor officials said they have solicited input on potential revisions from CIRPC; their on-site verification contractor; the U.S. Census Bureau, which is contracted to mail out the survey forms for Labor; and their regional offices. During our interviews, a Labor official said the agency would like to solicit input on proposed changes from survey respondents, but could not provide specifics. Although part of Labor's on-site verification process is to ask contractors questions about using the current form, Labor needs feedback on proposed changes to assess whether they will accomplish the goals of eliminating confusion and reducing errors.[Footnote 55] OMB guidance states that careful questionnaire design and pretesting can reduce measurement error and provide insights into how alternative wording can affect survey respondents' answers. Pretesting the new survey form with respondents to ensure changes achieve the desired results will be particularly important given that a Labor official said changing the form is a major undertaking. Labor officials did not have a specific time frame for implementing the new form because they said they are waiting for upgrades to the wage survey data system and their first priority is improving the online version of the form. Planned improvements to the online version include allowing respondents to save information rather than having to complete a survey before exiting. Seven stakeholders we interviewed agreed the ability to fill out the form online was important, but four of the seven were unaware it was already an option. Key Wage Determination Information Is Sometimes Confusing or Missing for Users: Labor's Davis-Bacon prevailing wage rates are publicly reported online at Wage Determinations Online for use by contractors and others to prepare bids for and pay workers on federal construction projects. [Footnote 56] While 6 of 27 stakeholders we interviewed said the general contractor provided the necessary wage information or they found the online wage determinations relatively easy to use, others reported problems. For example, while OMB and Labor guidance on data quality states that "influential" financial information provided by the agency should include a high level of transparency on data and methods, 15 stakeholders said there is a lack of transparency in the wage determinations because key information is not available or hard to find. In addition, both union and nonunion stakeholders said Labor's wage determination Web site should more clearly present information on the number of workers and wage rates used to calculate prevailing wages for each job classification. Labor currently makes some of this information available in a report known as a WD-22. The printed WD-22 provides, for each job classification, information on the final prevailing wage and fringe benefit rates, the total number of workers reported, and the method of rate calculation--for example, whether the rate was based on a majority or an average (see figure 7). A WD-22 is created for each state survey by construction type, but this information is not available on Labor's wage determination Web site. A senior Labor official said the WD-22 information is currently available upon request; though, the agency is considering posting it online along with other information used to determine wage rates. Figure 7: Labor's Printed WD-22 Report for a Texas Building Survey: [Refer to PDF for image: illustrated form] Wage Compilation - Texas, Building: El Paso: Depicted on the form: Job classification: Bricklayer. Recommended prevailing wage and fringe benefit rates: Wage: $13.45; Fringe: 0. Total number of workers reported: 26. Signals wage rate based on average: A. Signals wage rate issued with county-level data only: 0. Job classification: Caulker, Joint-Sealant. Recommended prevailing wage and fringe benefit rates: Wage: $10.00; Fringe: 0. Total number of workers reported: 6; Number of workers paid majority wage rate: 5. Signals wage rate based on majority: M. Source: GAO analysis of WD-22 excerpt from Labor’s Prevailing Wage Resource Book 2009. [End of figure] One contractor association representative said Labor's Web site does not explain the meaning of terms and codes, and a contractor said it is difficult to know which wage rates are in effect. Our review of the wage determinations posted online found some information confusing. Information provided about specific job classifications differs depending on whether the prevailing rate is nonunion or union. Nonunion rates are preceded with an "SU" designator, which is defined at the bottom of each wage determination as wage and fringe benefit rates that do not reflect collectively bargained rates (see figure 8). Union rates are preceded with four-letter designators, which are defined at the bottom of the wage determination as designations for unions whose rates have been determined to be prevailing. A prevailing wage resource book on another page of Labor's Web site further explains these designators. For example, it states the "SU" designator stands for "survey" and the accompanying date is the publication date of the survey used by Labor to set the rate. It also states that the date following the union designator is the effective date of the current collectively bargained rate. However, these explanations are not provided within the wage determination and, without understanding the date following the "SU" designator, users do not know when the survey that set posted rates was conducted. (See appendix V for the full wage determination.) Both contractor association and union representatives said more transparency about how prevailing wages are determined could potentially result in a better survey response because lack of understanding can deter stakeholders from participating. Figure 8: Excerpt from a Florida Wage Determination Published on Labor's Web Site: [Refer to PDF for image: illustrated form] Depicted on the form: General Decision Number: FL100104; 10/08/2010 (date updated during "roll-over" process and when wage determination modifications are issued); FL104. Superseded General Decision Number: FL20080104. State: Florida. Construction Type: Building. County: Alachua County in Florida. Building Construction Projects (does not include single family homes or apartments up to and including 4 stories). Modification Number: 0; Publication Date: 03/12/2010. Modification Number: 1; Publication Date: 03/19/2010. Modification Number: 2; Publication Date: 03/26/2010. Modification Number: 3; Publication Date: 05/14/2010. Modification Number: 4; Publication Date: 07/23/2010. Modification Number: 5; Publication Date: 08/06/2010. Modification Number: 6; Publication Date: 10/08/2010. ELEC[International/nations union] 1205[Local union] -004 [Internal number used by Labor for processing] 05/21/2010[Date of current negotiated rate]. Electrician: Rates: $23.48; Fringes: $6.46. SU[Survey designator] FL[State abbreviation] 2009[Year survey was published] 05/22/2009[Publication date of survey that set the rates following the "SU" designator] Cement Mason/Concrete Finisher: Rates: $17.69; Fringes: $1.83. Insulator - Pipe and pipewrapping: Rates: $13.13; Fringes: $3.03. In the listing above, the "SU" designation means that rates listed under the identifier do not reflect collectively bargained wage and fringe benefit rates. Other designations indicate unions whose rates have been determined to be prevailing. Source: GAO analysis of data from Labor’s Web site (www.wdol.gov) for Florida wage determination 104 as of October 13, 2010. [End of figure] Labor also changes the date at the top of a wage determination each calendar year in a "roll-over" process. Officials said the date is changed to inform users the posted wage rates are valid for the current year, but the wage rates contained in the determination are not necessarily updated. In the Florida example (see figure 8), the date at the top of the wage determination is October 8, 2010, but wage rates associated with the "SU," or survey, designator on the lower half of the page are from May 22, 2009, the publication date of the survey used to set those rates. A senior Labor official was not aware of users confusing the roll-over date on the wage determination with the survey publication date. However, OMB guidance states that when disseminating information products to users, key variables should be defined and the time period covered by the information and the date last updated should be provided. Not clearly explaining each of these dates within the wage determination reduces the transparency of when the last survey was conducted for an area, especially if many years have passed. Additionally, if the wage determination only contains union-prevailing rates, it does not contain any information about when the area was last surveyed. Finally, 9 of 27 stakeholders said missing wage rates are also a challenge. Specific job classifications may be missing from a wage determination if Labor received insufficient survey data. If job classifications are missing, contractors do not know what to bid on federal projects because they do not know what they will have to pay some workers, workers do not know what pay they will receive, and federal contracting agencies cannot accurately estimate costs. When a wage rate for a job classification is missing from the wage determination, it must be requested from Labor through a conformance process.[Footnote 57] While federal projects have contracting officers who typically request the conformance on behalf of the contactor, eight stakeholders said the contracting officers may not be familiar with the prevailing wages or the conformance process. Representatives from one national contractor association said the difficulty of bidding on projects when wage rates are missing, and then having to file a conformance request in order to know what to pay, can deter smaller contractors who might otherwise be interested in federal work. A Labor official said the rates issued via conformance requests--an average of over 3,000 per year were filed in fiscal years 2007, 2008, and 2009--are only good for the specific project on which they are issued and many are repeated requests for job classifications for workers who operate specific pieces of highway construction equipment. The best way to reduce conformance requests, the official said, is to conduct surveys that report wage rates for all job classifications. Some Stakeholders Are Not Aware of Labor's Outreach Efforts: The pre-survey briefing is one of Labor's primary outreach efforts to inform stakeholders about an upcoming survey. These briefings are conducted by regional office staff either before or at the start of a survey. A headquarters Labor official said regional offices notify state contractor associations and work through the Building & Construction Trades Department to notify unions about pre-survey briefings and ask them to pass the information along to their members. While the official said there is no required number of pre-survey briefings, regional office officials said they ranged from one briefing for two states to five briefings within one state for recent surveys depending on a state's size and characteristics. Officials said they generally hold separate briefings for unions and nonunion contractors/contractor associations. The presentation includes information on how wage and fringe benefit data are obtained and compiled, sufficiency requirements for issuing rates and wage determinations, and the process for filing conformances and wage determination appeals. A headquarters official said they are currently revising the presentation's information on how to fill out the survey form. Stakeholder awareness of the pre-survey briefings was mixed. In three states surveyed for building and heavy construction in either 2009 or 2010--Arizona, North Carolina, and West Virginia--all the union representatives we interviewed said they were aware of the pre-survey briefing and representatives from four of the six state contractor associations we interviewed said they were aware a briefing had been conducted.[Footnote 58] Of the 12 contractors we interviewed in Florida and New York who were last surveyed in 2005 and 2006, respectively, none were aware that a briefing had been conducted prior to the survey. Several regional office officials said the pre-survey briefings for unions generally have greater attendance than those for contractors. While one stakeholder said copies of the slides were provided at the briefing, a Labor headquarters official said the information is not available online for those who are unable to attend in person. Seven of 27 stakeholders indicated that alternative approaches, such as webinars or audioconferences, might be helpful ways to reach additional contractors.[Footnote 59] CIRPC officials said more outreach by Labor could improve the accuracy of the surveys because contractors would better understand why and how the surveys are conducted, thereby encouraging more to participate. They said they previously recommended that Labor wage analysts call contractors prior to survey distribution to make them aware of the survey and to assure them their submitted data would be protected. OMB guidance states that sending a letter in advance of a survey to alert respondents can improve response rates. A senior Labor official said they are conducting pre-survey briefings instead of calling respondents in advance. Conclusions: For more than a decade, reviews of the Davis-Bacon wage survey have highlighted methodological problems in the determination of wages paid to workers on federally funded construction projects. In response to those criticisms, Labor has improved its process, most recently seeking out new data sources for some construction types and adjusting the data collection and processing time frames. Yet without clear tracking of key survey dates and the time spent in various processing activities, Labor cannot assess if its changes are improving survey timeliness and thus the accuracy of published wage rates. Additionally, these efforts do not effectively address some key issues with how data are collected. Because Labor has not conducted checks over the past several years on the representativeness of the data it receives, it cannot have high confidence its results accurately reflect prevailing wages, no matter how diligently its staff work to clarify and verify submitted data. If the resultant prevailing wage rates are too high, they potentially cost the federal government and taxpayers more for publicly funded construction projects or, if too low, they cost workers in compensation. While Labor officials rightly used experience and corporate knowledge in designing recent changes to survey methodology, they did not enlist objective survey expertise to ensure methods were sound and in accordance with best practices. Survey methodology that does not follow best practices lowers confidence in the process and puts participation by private contractors at risk. Labor's regulatory goal to issue wage rates at the county level may also limit its ability to improve survey representativeness and timeliness. Labor often must combine data from multiple counties to meet its own relatively low sufficiency standards to publish wage rates for specific job classifications which, in the end, may reflect the wages for as few as three employees from two contractors for an entire state. The statutory requirement to issue prevailing wages by "civil subdivision of the state" limits Labor's ability to account for relevant regional markets that cross county or state boundaries or to tap into data based on other geographic groupings. Use of other data sources to augment Davis-Bacon survey data could shorten the time needed to publish wage rates and reduce the number of conformances that contractors must file for missing wage rates. Given the voluntary nature of the survey, participants who take the time to respond should have confidence their information will be considered in determining prevailing wages. They should also be able to understand how their information is used. Increased transparency in how the wage rates are calculated and improved clarity in published wage determinations would provide stakeholders assurance the wage rates are accurate and encourage greater participation of the construction employer community. Matter for Congressional Consideration: To improve the quality of Labor's Davis-Bacon wage survey data, Congress may wish to consider amending the language of the Davis-Bacon Act to allow Labor to use wage data from geographic groupings other than civil subdivisions of states, such as metropolitan statistical areas or Bureau of Economic Analysis' economic areas. Recommendations for Executive Action: To improve the quality and timeliness of Labor's Davis-Bacon wage surveys, we recommend that the Secretary of Labor direct the Wage and Hour Division to enlist the National Academies, or another independent statistical organization, to evaluate and provide objective advice on the survey, including its methods and design; the potential for conducting a sample survey instead of a census survey; the collection, processing, tracking, and analysis of data; and promotion of survey awareness. To improve the transparency of wage determinations while maintaining the confidentiality of specific survey respondents, we recommend that the Secretary of Labor direct the Wage and Hour Division to publicly provide additional information on the data used to calculate its Davis- Bacon wage rates, such as the number and wages of workers included in each wage rate calculation, and to clearly communicate the meaning of various dates and codes used in wage determinations in the same place the prevailing wage rates are posted. Agency Comments and Our Evaluation: We provided a draft of this report to Labor for review and comment. The agency provided written comments, which are reproduced in appendix VI. Labor agreed with our recommendation to improve the transparency of the wage determinations and indicated it is taking steps to do so. However, the agency said our recommendation to obtain objective expert advice on its survey design and methodology may be premature because additional changes are currently being implemented or will be implemented based on a 2004 review of the program by McGraw-Hill Construction Analytics. The McGraw-Hill review was a process evaluation that assessed many aspects of the wage survey; however, Labor officials did not indicate during our interviews that the results of that evaluation were serving as the foundation for their recent changes nor was the evaluation referred to in documentation Labor provided regarding its recent changes. Moreover, the McGraw-Hill report did not address certain issues related to the survey's design and methodology. Therefore, we continue to believe that Labor should have an independent statistical organization provide advice on survey methods for the following reasons: * Labor cites examples of improvements to its processes and information technology systems so that surveys can be completed and published in a more timely manner. We also cited many of these data collection and processing changes in our report along with the agency's expected reduction in processing times for highway, building, and heavy surveys. The survey timelines, which we used to assess whether surveys conducted under new processes were on schedule, were provided to us by the agency and included reductions in and elimination of various survey steps. Yet according to those agency timelines, many of the surveys were behind schedule. Labor commented it has reduced the time to publish survey results for building and heavy construction from several years to an average of 2 years. However, we believe it may face challenges staying on schedule if it cannot more accurately predict how many survey forms it will receive and the time required to process them. Possibilities to better predict the number of survey responses, such as statistical sampling rather than the current census survey, could be explored with survey experts. * Labor also noted, as we did in our report, that it is again working to calculate response rates and we believe this is a step in the right direction. However, only calculating response rates will not ensure that the data Labor is using to calculate prevailing wages are truly representative of the wages being paid in a particular area. If a response rate is low--some wage rates are calculated on as few as three workers--then Labor must also analyze nonrespondents to ensure that those who received a survey but did not respond do not significantly differ from those who responded. Survey expertise could assist with this critical data quality check to help ensure prevailing wages are representative of wages actually paid to workers. * Labor commented that the current survey form was not recently redesigned, but is a scannable version of the form that was last updated in 2004. We adjusted our report language accordingly. The agency also noted that errors on wage survey forms typically result from errors in the information provided by survey respondents rather than errors made by Wage and Hour Division employees. We agree; however, we believe the fact that respondents continue to make some of the same errors in completing the wage survey form that were identified by the Labor OIG in 2004 is a concern. Labor did not pretest the current form with survey respondents to ensure clarity, which could partially explain why contractors and interested parties made errors. A professional survey methodologist could develop a pretesting plan to address issues that affect the quality of the survey data, such as respondent comprehension, retrieval, judgment, and response formulation. We believe it is critical for Labor to obtain expert methodological advice because this would allow the agency to make course corrections before time and money are spent implementing new procedures that may increase the speed of processing data, but not sufficiently address its quality. While Labor indicated the cost of contracting for an expert review is a concern, not ensuring the quality and representativeness of the data can be costly in other ways: the federal government could pay more for construction than it needs to or workers may earn less than they should. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days from its issue date. At that time, we will send copies of this report to the Secretary of Labor, relevant congressional committees, and other interested parties. The report will also be available at no charge on the GAO Web site at [hyperlink, http://www.gao.gov]. If you or your staff have questions about this report, please contact me at (202) 512-7215 or sherrilla@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix VII. Sincerely yours, Signed by: Andrew Sherrill, Director: Education, Workforce, and Income Security: [End of section] Appendix I: Objectives, Scope, and Methodology: Our review examined (1) the extent to which the Department of Labor (Labor) has addressed concerns regarding the quality of the Davis- Bacon wage determination process and (2) the additional issues identified by stakeholders regarding the wage determination process. To address these objectives, we: * reviewed key documents, including past GAO and Department of Labor Office of Inspector General (OIG) reviews of the program, agency documents on recent changes to the wage survey process, and relevant federal laws and regulations; * interviewed agency officials and representatives from organizations with whom the agency contracts some aspects of the survey process; * analyzed (1) data from Labor's Automated Survey Data System (ASDS), Wage Determination Generation System (WDGS), and the Davis-Bacon survey schedule Web site [hyperlink, http://www.dol.gov/whd/programs/dbra/schedule.htm]; (2) reports produced by Labor's contracted accounting firm for on-site verification of submitted payroll records; and (3) Labor's conformance logs for fiscal years 2007 through 2009; * conducted site visits to three of Labor's five regional offices that conduct Davis-Bacon wage surveys, as well as to the Construction Industry Research and Policy Center (CIRPC), which is contracted to assist Labor with the wage survey process; * interviewed approximately 30 stakeholders, including representatives from academia, contractor associations, and unions, as well as individual contractors and performed a content analysis of their comments; and: * attended a Labor prevailing wage conference. We conducted this performance audit from September 2009 through March 2011 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Review of Key Documentation and Interviews with Agency Officials: To evaluate how Labor has addressed past concerns with the quality of the Davis-Bacon wage determination process, we reviewed past reports, reviewed key agency documents, and interviewed Labor officials. We reviewed two Labor OIG reports and their associated recommendations,[Footnote 60] as well as our own previous work. [Footnote 61] In addition, we reviewed agency correspondence with Congress and the Office of Management and Budget (OMB) describing Labor's changes to the wage determination process based on past program audits, the effectiveness of those changes, and planned future changes.[Footnote 62] To assess recent changes made to the wage survey process and their expected outcomes, we interviewed officials and reviewed agency documents, such as the Davis-Bacon manual of operations and Labor's revised timelines for building, heavy, and highway surveys starting in 2009. Using Labor's revised timelines, we calculated the expected reduction in the amount of time from the start of each survey to publication of wage rates. To assess whether Labor's surveys under the new processes were on schedule, we reviewed an ASDS Individual Time Tracking Report by Activity/Survey for October 1, 2009, through September 1, 2010, that provided the number of staff hours logged in each survey activity for the pilot building and heavy surveys and highway surveys started under the new processes.[Footnote 63] We then compared the last activity in which staff hours had been logged for each survey with its expected activity based on the date the regional office entered the survey into ASDS and Labor's new timelines. We could not calculate the exact number of days surveys were ahead of or behind schedule because Labor did not have a report that reliably recorded the date a survey moved from one activity to the next. Additionally, for one state building and heavy survey and three state highway surveys, we could not calculate the actual timelines because the dates in the data provided by Labor were out of sequence. Labor officials provided inconsistent guidance on which activity in their timelines reflected the actual start of a survey; however, for various reasons, we used the date the survey was recorded as being entered into the database for our analysis of whether the surveys were on schedule. During our review, a senior Labor official indicated the appropriate survey start date was the date the survey was entered into ASDS by regional office officials. Toward the end of our review, the official indicated the correct start date was the date surveys were first mailed to contractors or interested parties because each region had its own method for when it entered surveys into ASDS. For example, some regions entered surveys when they planned them while others entered surveys when they ordered Dodge Reports. We believe using the date surveys were first mailed as the start date would exclude certain key activities on Labor's survey timeline, such as ordering, receiving, and cleaning the Dodge Report data for building and heavy surveys and inputting interested party lists for highway surveys. Nonetheless, we conducted an additional timeliness analysis using alternative start dates based on Labor's concerns. Given that Labor officials were concerned the regional offices may enter building and heavy surveys into ASDS before actually starting them, we used the date the Dodge data were requested, which is the second step in the new process. For the building and heavy surveys we reviewed, none of them changed status based on the alternative start date. In other words, all were still behind schedule. For highway surveys, we used the date surveys were first mailed to interested parties as the start date for the alternative analysis. For the highway surveys we reviewed, only one changed status from behind schedule to ahead of schedule. Therefore, based on the limited changes to our findings from using alternative start dates, as well as the fact that the alternative start dates exclude parts of the survey process on which Labor had been working to improve timeliness and for which staff had logged hours, we decided to conduct our analysis using the original date provided by Labor (the date the regional offices entered the survey into ASDS). To assess the adequacy of Labor's current wage survey methodology we compared it with survey guidance published by OMB and Labor.[Footnote 64] Analysis of Survey Data: We used data from ASDS to evaluate the geographic level at which rates were issued and the number of workers used to issue rates. For both analyses, we used data from four surveys--Florida 2005, Maryland 2005, Tennessee 2006, and West Texas Metropolitan 2006--that were issued in 2009 or 2010. We selected these surveys because they were recently published and represented geographic diversity, to the extent possible, in terms of the Labor regional offices that conducted the surveys. The data from the surveys we reviewed included the following construction types: Florida--building, heavy, highway, and residential; Maryland--building, heavy, and residential; Tennessee-- building, heavy, highway, and residential; and West Texas Metropolitan--building and residential. The survey results included metropolitan and rural rates for all construction types with the exception of the Maryland heavy construction type and the West Texas survey, which only included metropolitan rates. To evaluate the geographic level at which wage rates were issued, we analyzed, for each survey in our review, the "calculation basis" field on Labor's WD-22 form, which indicates whether the wage rate for each job classification was determined based on county-level data, multi- county data, or statewide data.[Footnote 65] We were unable to determine the geographic level for rates that had been combined in the final WD-22 so we reported them separately. Regional office officials said they may combine rates from counties with the exact same wage and fringe benefit data in the final WD-22. However, the rates being combined may have been calculated at different geographic levels--for example, one county's rates may have been calculated at the group level while another county's rates may have been calculated at the supergroup level.[Footnote 66] Because the geographic level at which rates for each combined county were calculated is not reported on the WD-22, we reported the percentage of these rates separately. We analyzed geographic levels for key job classifications only because nonkey job classifications cannot be issued at the supergroup or state level. Key job classifications are those determined by Labor to be necessary for one or more of the four construction types, as follows: * Building Construction: bricklayer, boilermaker, carpenter, cement mason, electrician, heat and frost insulators/asbestos workers/pipe insulators, iron worker, laborer-common, painter, pipefitter, plumber, power equipment operator, roofer, sheet metal worker, tile setter, and truck driver. * Heavy Construction and Highway Construction: carpenter, cement mason, electrician, iron worker, laborer-common, painter, power equipment operator, and truck driver. * Residential Construction: bricklayer, carpenter, cement mason, electrician, iron worker, laborer-common, painter, plumber, power equipment operator, roofer, sheet metal worker, and truck driver. Table 1 provides the percentage of wage rates issued at each geographic level by construction type and metropolitan or rural designation for the four surveys we reviewed. Table 1: Percentage of Key Job Classification Wage Rates Issued at Each Geographic Level by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: County: Building: Metro: 16%; Building: Rural: 5%; Heavy: Metro: 10%; Heavy: Rural: 7%; Highway: Metro: 11%; Highway: Rural: 7%; Residential: Metro: 9%; Residential: Rural: 6%. Group: Building: Metro: 22%; Building: Rural: 39%; Heavy: Metro: 19%; Heavy: Rural: 29%; Highway: Metro: 11%; Highway: Rural: 39%; Residential: Metro: 14%; Residential: Rural: 21%. Supergroup: Building: Metro: 19%; Building: Rural: 26%; Heavy: Metro: 19%; Heavy: Rural: 9%; Highway: Metro: 18%; Highway: Rural: 19%; Residential: Metro: 25%; Residential: Rural: 12%. State: Building: Metro: 42%; Building: Rural: 21%; Heavy: Metro: 43%; Heavy: Rural: 25%; Highway: Metro: 55%; Highway: Rural: 15%; Residential: Metro: 42%; Residential: Rural: 23%. Geographic level not available: Building: Metro: 1%; Building: Rural: 9%; Heavy: Metro: 9%; Heavy: Rural: 30%; Highway: Metro: 4%; Highway: Rural: 20%; Residential: Metro: 9%; Residential: Rural: 38%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in 2009 or 2010. Note: Percentages may not equal 100 due to rounding. [End of table] We also used WD-22 data to determine the number of workers used to calculate wage rates for all key job classifications for the four surveys in our review. Using the "total number reported" column in WD- 22 reports, we calculated the number of workers whose wage rates were included in each wage rate calculation for key job classifications. We reported the data by quartiles with the exception of the "3 workers" category, which we broke out separately because it is the minimum number of workers for which Labor must receive data in order to issue a wage rate for a job classification. Table 2 provides the percentage of key job classification rates issued by number of workers, construction type, and metropolitan or rural designation for the four surveys we reviewed. Table 2: Percentage of Key Job Classification Wage Rates Issued Based on Number of Workers by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: 3 workers; Building: Metro: 75%; Building: Rural: 25%; Heavy: Metro: 79%; Heavy: Rural: 21%; Highway: Metro: 65%; Highway: Rural: 35%; Residential: Metro: 52%; Residential: Rural: 48%. 4-6 workers; Building: Metro: 71%; Building: Rural: 29%; Heavy: Metro: 88%; Heavy: Rural: 12%; Highway: Metro: 75%; Highway: Rural: 25%; Residential: Metro: 82%; Residential: Rural: 18%. 7-12 workers; Building: Metro: 65%; Building: Rural: 35%; Heavy: Metro: 88%; Heavy: Rural: 12%; Highway: Metro: 78%; Highway: Rural: 22%; Residential: Metro: 71%; Residential: Rural: 29%. 13-28 workers; Building: Metro: 75%; Building: Rural: 25%; Heavy: Metro: 87%; Heavy: Rural: 13%; Highway: Metro: 80%; Highway: Rural: 20%; Residential: Metro: 85%; Residential: Rural: 15%. 29 or more workers; Building: Metro: 72%; Building: Rural: 28%; Heavy: Metro: 100%; Heavy: Rural: 0%; Highway: Metro: 85%; Highway: Rural: 15%; Residential: Metro: 92%; Residential: Rural: 8%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in 2009 or 2010. [End of table] Finally, we used WD-22 data to determine the percentage of wage rates that included federal data. We calculated this percentage for the building and residential construction types for the surveys in our review because Labor uses federal data for these construction types only when it has insufficient survey data, whereas federal data are used in all highway and heavy surveys. Table 3 provides the percentage of key job classification wage rates using federal data by construction type and metropolitan or rural designation for the four surveys we reviewed. Table 3: Percentage of Key Job Classification Wage Rates Using Federal Data by Construction Type and Metropolitan or Rural Designation, for Four Surveys Reviewed: Used federal data; Building: Metro: 24%; Building: Rural: 20%; Residential: Metro: 61%; Residential: Rural: 95%. Did not use federal data; Building: Metro: 76%; Building: Rural: 80%; Residential: Metro: 39%; Residential: Rural: 5%. Source: GAO analysis of Labor data from Florida, Maryland, Tennessee, and West Texas Metropolitan surveys published in 2009 or 2010. [End of table] To determine the age of wage rates, we used WDGS data on published wage rates provided by Labor officials on November 12, 2010. We analyzed the age of wage rates for building, heavy, and highway construction because Labor considered only those construction types in its fiscal year 2010 performance goal.[Footnote 67] We analyzed the age of wage rates in two ways: first, combining nonunion-and union- prevailing wage rates together, as Labor does, and then separately to identify any trends by type of rate. To determine the age of data used to calculate prevailing wage rates for the 22 open surveys that accumulated since Labor began conducting statewide surveys, we analyzed survey time frames and cutoff dates from Labor's Davis-Bacon and Related Acts survey schedule Web site [hyperlink, http://www.dol.gov/whd/programs/dbra/schedule.htm] and interviewed Labor officials. To assess the number of wage survey forms, or WD-10s, that had errors and the types of errors that most commonly occurred, we analyzed on- site verification reports prepared by Labor's contracted accounting firm for the four states in our review. We analyzed the verification reports to determine what percentage of wage survey forms that were verified had errors and what type of errors occurred. To identify and categorize the errors, we recorded if the accounting firm marked an error in the following fields: project value, construction type, additional trade/classification, employee classification, work performed, paid under collective bargaining agreement (CBA), number of employees, peak week, hourly rate, fringe benefit, health and welfare, pension, holiday and vacation, apprentice training, and other. We counted a wage survey form as having multiple errors if it had an error in more than one category. To determine the average number of conformance requests filed for missing classifications in fiscal years 2007 through 2009, we used the "tracking number" field in Labor's conformance request log. We counted the number of requests with distinct tracking numbers, excluding entries that did not have tracking numbers, and then calculated the average over the 3-year period. To assess the reliability of the data we used in our analyses, we performed the following steps: (1) reviewed pertinent system and process documentation, (2) interviewed agency officials knowledgeable about the data and system during each regional office site visit, and (3) performed electronic testing of required data fields. We found the data we reviewed to be reliable for our purposes. Regional Office Site Visits: To obtain information from staff who clarify and analyze survey information, we conducted site visits to three of the five Labor regional offices that process Davis-Bacon wage surveys--Northeast region (Philadelphia, Pennsylvania); Southeast region (Atlanta, Georgia); and Southwest region (Dallas, Texas)--as well as CIRPC at the University of Tennessee.[Footnote 68] At each regional office, we interviewed the director of enforcement, the regional wage specialist, the senior wage analyst, and wage analysts. At CIRPC, we interviewed the associate directors, the senior wage analyst equivalent, and wage analysts. Also, to gain a thorough understanding of how wage analysts process survey data and document decisions, we interviewed staff at each regional office about ASDS. We selected our site visit locations based on the fact that Labor headquarters officials said these regional offices were currently conducting surveys using new processes. Additionally, we visited CIRPC to determine how contractors are selected for survey participation and on-site verification, and how CIRPC provides support to the regional offices in implementing the new survey processes. Interviews with Stakeholders: To determine what additional issues stakeholders may have with the wage determination process, we initially explored surveying contractors and union officials in states where Labor had recently conducted a wage survey. We believed it was important for us to survey contractors who had recently received a wage survey from Labor so they could recall their experience of responding to the wage survey or their reasons for not responding. However, Labor officials had concerns about us surveying contractors in states where Labor had completed wage survey data collection, but was still in the process of contacting contractors for data clarification and verification. Labor officials believed contractors might get confused if they received requests for information from more than one agency and were concerned our activities might affect their efforts. We agreed with these concerns. Therefore, instead of surveying contractors, we opted to conduct semi-structured interviews with a wide variety of Davis-Bacon stakeholders. Also, in order to solicit opinions directly from contractors but not interfere with Labor's ongoing efforts, we interviewed a small number of individual contractors in states that had been surveyed less recently but where the results of those wage surveys had been published. Given that it had been a few years since Labor sent wage survey forms to these contractors, we believed we would obtain better information through personal interviews than a survey. We conducted semi-structured interviews with approximately 30 representatives from academia, contractor associations, unions, and individual contractors. Our semi-structured interview protocol allowed us to ask questions of numerous organizations and individuals, offering each interviewee the opportunity to respond to the same general set of questions, but also allowed for flexibility in asking follow-up questions and, in limited circumstances, for the omission of questions when appropriate. For example, we did not ask representatives from academia about filling out the survey form or attending pre-survey briefings because they would typically not be involved in these activities. In our findings, we noted cases in which we did not ask all stakeholders a particular question. To select representatives from academia, we conducted a literature review to identify studies that reviewed or evaluated the Davis-Bacon wage survey process. To obtain opinions from both unionized and nonunionized contractors, we interviewed representatives from the national organizations of the Associated Builders and Contractors, Inc. (ABC) and the Associated General Contractors of America (AGC). To obtain views from construction unions, we interviewed representatives from the AFL-CIO and the International Brotherhood of Electrical Workers (IBEW). We selected IBEW because it has one of the largest memberships among construction industry unions and electricians are considered a key class for all four of Labor's construction types. To obtain a state-level perspective from contractors' associations and unions, we interviewed representatives from state ABC and AGC chapters, as well as IBEW locals in Arizona, North Carolina, and West Virginia. We chose these three states because they had been surveyed in 2009 or 2010 by different Labor regional offices. In addition, because Arizona, North Carolina, and West Virginia have low to medium levels of workers represented by unions, according to the Bureau of Labor Statistics (BLS), we interviewed representatives from ABC and AGC in New York, the state with the highest level of unionization. Footnote 69] We also interviewed individual contractors in New York and Florida. We chose New York and Florida because they had been surveyed fairly recently and represented diversity in geography and the percentage of all workers represented by unions.[Footnote 70] To select contractors, we requested Labor data including the lists of contacts who had been sent wage survey forms and who had returned them. Then, to the extent possible, we matched the data using the contact identification field to determine which contacts had responded or not responded. In each state, we identified the counties with the highest number of respondents because there were fewer respondents than nonrespondents. We selected certain ZIP codes within each selected county based on the highest concentration of respondents, as well as site visit logistics. We then ordered the list of respondents and nonrespondents by ZIP code and called contractors asking them to meet with us. If we were unable to reach a contractor or if a contractor declined, we moved to the next contractor on the list and continued until we had a mix of respondents and nonrespondents who agreed to be interviewed. We conducted a content analysis on the information gathered through the stakeholder interviews. Interview responses and comments were categorized by an analyst to identify common themes. A pretest of the themes was reviewed by the engagement's methodologist before all comments were categorized. The categorization of the comments was then independently checked, and agreed upon, by another analyst for verification purposes. While we selected our stakeholders to include a wide variety of positions, the opinions expressed are specific to those we interviewed and are not generalizable. Prevailing Wage Conference: We attended Davis-Bacon-related sessions of Labor's November 2010 prevailing wage conference in Cleveland, Ohio, to obtain additional stakeholder perspectives on the wage determination process and use of published wage determinations through observation of Labor's presentations and question and answer sessions. [End of section] Appendix II Wage Survey Form (WD-10): [Refer to PDF for image: wage survey form] Form WD-10: Davis-Bacon Wage Survey: Report of Construction Contractor's Wage Rates: U.S. Department of Labor: Employment Standards Administration: Wage and Hour Division: 1. Name Of Contractor/Subcontractor: Address: City: State: Zip: Phone: Extension: Fax: 2. Submitter information: Title: Organization: Phone: Extension: Fax: Email address: 3. Full Name of Project: Address: City: State: County: Name of General/Prime Contractor: Instructions - Please enter the information in the white boxes and fill in the circles as appropriate. You can either hand print the information in blue or black ink, or use a typewriter or printer. Detailed instruction for completing this form (or obtaining additional copies) as well as definitions of many of the terms used on this form are found or a separate instruction page. We estimate that it will take an average of 20 minutes to complete this collection of information, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. If you have any comments regarding the burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden, send them to: U.S. Department of Labor, Wage and Hour Division Administrator, Room S- 3502, 200 Constitution Avenue NW, Washington, DC 20210. Note: This form is used by the U.S. Department of Labor to determine the locally prevailing wage rates under the Davis-Bacon Act. The submission of wage data is encouraged but is voluntary. This is an optional form provided to ensure consistency in submission of wage data. Respondents may use an alternate form if all the information requested is included. The identity of the Respondent will be kept confidential to the maximum extend possible under existing law. Persons are not required to respond to this collection of information unless it displays a currently valid OMB control number. 4. Indicate if project is subject to a Federal (Davis-Bacon) or state wage determination: Federal: State: Neither: 5. Indicate: I am the: General/Prime Contractor: Subcontractor: 5A. Please provide a list, on the enclosed form, of any subcontractors you used on the project, including addresses and phone numbers: The list is being returned with this form: The list was provided earlier: There are no subcontractors: 5B. For the project being reported on this form, state the date the work: Began: Ended: Estimated/Actual Project Value: 5C. If you are a subcontractor for the project being reported indicate the date your work: Began: Ended: Estimated/Actual Project Value: 6. Please fill in the circle indicating the type of construction for the project being reported and all relevant descriptors. If the project has more than one type of construction, please mark the additional type: Apartment building: Bicycle path: Bridge over navigable water: Bridge (any other type): Dormitory: Hospital: Motel/Hotel: Nursing/Assisted living facility: Office/Commercial building: Paving: Parking lot: Playground: Residential: Road/Street/Highway/Drive: School: Site Preparation: Treatment Plant: Water/Sewer: Other: If you selected apartment, nursing facility or residential: Number of stories: Kitchen in each unit: Bath in each unit: 7. Classification and Fringe Benefit Information. In the question below, CBA stands for Collective Bargaining Agreement. In the five benefit-related columns, please describe the benefits (if any) for each classification, and also tell us how they are paid. If the benefit is paid out periodically, tell us how much you pay and how frequently you pay it, using a single letter abbreviation. Use 'H' for hourly, 'D' for daily, 'W' for weekly, and 'Y' for yearly. If the benefit is paid as a percentage of the hourly rate, check the appropriate box, then tell us the percentage using the boxes below the checkboxes. Regarding the Vacation and Holiday and additional benefit columns, if appropriate, tell us how many days are paid annually. If you only supplied building materials, and no employees worked on the project, then fill in the circle below. You may skip the rest of the question and sign and date the form. Only Supplied Materials: Classification: Peak Week Ending Date: Hourly Rate: Type Of Work Performed: Number Of Employees: Paid Under A CBA? Health and Welfare: $ per employee, per: % of hourly rate: Pension (401K, etc.): $ per employee, per: % of hourly rate: Apprentice training: $ per employee, per: % of hourly rate: Vacation and Holiday: $ per employee, per: % of hourly rate: # days per year: Additional fringe: $ per employee, per: % of hourly rate: # days per year: 8. Comments or remarks: Description of any additional fringe (see last column of item 7): Your signature: Note: The willful falsification of any submitted information may result in civil or criminal prosecution. See 18 U.S.C. 1001. Date: Instructions for the WD-10: Davis--Bacon Wage Survey: Report of Construction Contractor-s Wage Rates: Information about Davis-Bacon Wage Surveys, including dates of current and future surveys, may be obtained at the Davis-Bacon and related Acts (DBRA) web site at [hyperlink, http://www.dol.gov/whd/programs/dbra/index.htm]. Instructions: * Use blue or black ink. * Hand-print letters/numbers. * Use one block for each letter, number, period, or space. If you use a typewriter or printer to complete this form, ignore the block spacing. * Fill in circles completely. * Use one WD-10 form for each construction project. This form is machine readable, and should not be copied. For additional forms, please contact (1-866-487-9243), or fill out and submit your forms electronically using the following site on the World Wide Web: [hyperlink, http://www.dol.gov/whd/programs/dbrajwd-10.htm]. FORM SIDE 1: Sections 1 and 2 -- Contractor and Submitter Information: 1. Fill in with information about your company. 2. Fill in with information about the submitter of the form. Sections 3, 4, 5, and 6 -- Project Information: 3. Fill in information about the construction project your company worked on and the project's location and description. 4. Fill in one circle to identify if the project was subject to a federal or state wage determination. 5. Fill in one circle to identify yourself as either the general/prime contractor or a subcontractor. 5A. Indicate if you had no subcontractors, or if you did, then indicate whether you are enclosing a list of subcontractors along with the WD-10 form, or if you submitted a list earlier. 5B. If you were the prime/general contractor, provide the date any work began on this project, the date the project ended (indicate if actual or estimated date), and the total project value. 5C. If you were a subcontractor, provide the date your work started and ended (indicate if actual or estimated date) and the subcontract value. 6. Mark the type of construction project your company worked on. If none of the construction types matches your project, fill in the circle next to OTHER, and indicate the type of construction in the blocks. If you selected Apartment Building, Nursing/assisted Living Facilities, Or Residential, indicate the number of stories, and fill in the circle if there was a kitchen and/or a bath in each unit. Form Side 2: Section 7 -- Classification and Fringe Benefits: * If you only supplied materials, and no employees worked on the project, then fill in the circle marked "Only Supplied Materials," skip the rest of section 7, and sign and date the form. * The remainder of section 7 requests multiple types of information per classification. Fill in each item as defined and described as follows: - Classification(s) are the position titles of jobs within your company (e.g., Carpenter, Electrician, Laborer, Crane Operator, etc.). Fill in one classification per line. If the workers in a classification are paid more than one hourly rate or different fringe benefits, please list them on separate lines. If more than 6 classifications and wage rates need to be listed for a project, report the additional classifications and wage rates on a new WO-10. On the new WD-10 fill out only Sections 1, 3, and 7: Guam Survey Respondents Only: * List H2B Visa Workers Separately From Other Reported Workers. * Identify H2B Workers By An "H2" After The Classification Title. Example: H2B/Visa Carpenter: Classification: Carpenter H2. Non-H2B Carpenter: Classification: Carpenter. * Type of Work Performed - Explain the type of work that each classification performs (e.g., Laborer: landscape, unskilled, pipelayer; Carpenter: carpentry, drywall; Operator: backhoe, etc.). Section 7 -- Classification and Fringe Benefits (continued) : * Peak Week Ending Date is the week you had the largest number of employees in a classification. * Number of Employees is the largest number of employees working in this classification on this project. * Hourly Rate is the dollar amount you paid employees per hour working in this classification. * CBA -- If the employee is paid under a Collective Bargaining Agreement, fill in the circle that represents Yes, otherwise fill in the circle that represents No. * Fringe Benefits are paid in addition to the hourly rate. Report only the costs or contributions incurred by your company, NOT the employees. Do not include costs paid by the employer that are required by either Federal, State, or local law such as worker's compensation or unemployment insurance. Fill out the information under each fringe benefit that applies. - Health & Welfare -- Medical or hospital care, or insurance to provide such care, life insurance, long- or short-term disability, sickness, or accident insurance. - Pension (401K, etc.) -- Retirement/401K, defined contribution plans (including savings and thrift, deferred profit sharing and money purchase pension), annuity cost, or cost of insurance to provide such a benefit. - Apprentice Training -- Defrayment of the cost of apprenticeship or similar training programs. - Vacation & Holiday -- The payment of compensation for holidays and vacation. - Additional Fringe -- If you are not sure of the category of the fringe benefit(s), enter the rate information in the column, and specify the fringe type in the "Description of Any Additional Fringe" field at the bottom of the form. Fringe benefits can be paid by a straight dollar amount, or by a percentage of the basic hourly rate. Indicate the cost or contribution your company paid to this classification during the peak week of this project. If the fringe benefits were paid by a straight dollar amount: Dollars ($) per Employee (EMP.) per: * Mark the circle before $ per EMP. per: * Fill in the dollar value in the blocks provided. Include the decimal position when you fill in the dollar amount. Do not include the $ sign. (Example: 1.50 for one dollar and fifty cents.) * Indicate how often this dollar value was paid in the block following $ per EMP. per with the values as follows: H for Hourly, D for daily, W or weekly, M for monthly, and A for annually/yearly. Example- If an employee was provided a straight dollar amount of $1.50 on a weekly basis for health and welfare: Item 8 -- Comments or Remarks and Signature: * Comments or remarks -- Provide comments or additional information. * Signature -- Submitter must sign and date the form. Key Terms: * Apprentice -- A person employed and registered in a bonafide apprenticeship program. (If these Apprentices/Trainees are in a formal program approved by the U.S. Dept. of Labor, Bureau of Apprenticeship and Training (BAT), or a state apprenticeship agency recognized by BAT, then information regarding wages and fringe benefits need not be provided.) * Helper -- A person that helps or assists and whose duties are distinct from the journey level class and laborer. * General/Prime Contractor -- The principal contractor on the project. * Subcontractor -- A contractor working on the project responsible for specific work but not the overall project. You are not a subcontractor for purposes of this survey if you supplied only materials. * Subcontractor List -- A machine--readable form for reporting the names and addresses of any subcontractors used by the contractor/subcontractor on the project being reported. * Trainee -- A person registered in a construction occupation program. [End of section] Appendix III: Labor's Wage Determination and Appeals Process under the Davis-Bacon Act: The Davis-Bacon Act requires that workers employed on federal construction contracts valued in excess of $2,000 be paid, at a minimum, wages and fringe benefits that the Secretary of Labor determines to be prevailing for corresponding classes of workers employed on public and private projects that are similar in character to the contract work in the civil subdivision of the state where the construction takes place.[Footnote 71] To determine the prevailing wages and fringe benefits in various areas throughout the United States, Labor's Wage and Hour Division periodically surveys wages and fringe benefits paid to workers in four basic types of construction: building, residential, highway, and heavy.[Footnote 72],[Footnote 73] Labor collects data through statewide surveys, except in large states, such as Texas and California. Labor's regulations state that the county will normally be the civil subdivision at which a prevailing wage is determined, although Labor may consider wages paid on similar construction in surrounding counties if it is determined there has not been sufficient similar construction activity within the given area in the past year. [Footnote 74] Data from projects in metropolitan counties are considered separately from those in rural counties. If similar construction in surrounding counties, or in the state, is not sufficient, Labor may consider wages paid on projects completed more than 1 year prior to the start of a survey. Wage rates are issued for a series of job classifications in each of the four basic types of construction, so each wage determination requires the calculation of prevailing wages for many different trades, such as electrician, plumber, and carpenter. Labor's wage determination process consists of five basic stages: 1. Planning and scheduling surveys to collect data on wages and fringe benefits in similar job classifications on comparable construction projects. 2. Conducting surveys of employers and interested parties, such as representatives of unions or contractor associations. 3. Clarifying and analyzing respondents' data. 4. Issuing the wage determinations.[Footnote 75] 5. Reconsideration and review of wage determinations through an appeals process. Stage 1: Planning and Scheduling Surveys: Labor attempts to survey the complete "universe" of relevant construction contractors active within a particular area during a specific period of time. Labor schedules surveys by identifying those areas and construction types most in need of a survey, based on criteria that include: * age of the most recent survey;[Footnote 76] * volume of federal construction in the area; * requests or complaints from interested parties, such as state and county agencies, unions, and contractor associations; and: * evidence that wage rates in a region have changed. Labor uses two management tools, the Regional Survey Planning Report and the Uniform Survey Planning Procedure, to help prioritize planned surveys. The Regional Survey Planning Report is provided by CIRPC at the University of Tennessee and contains information about construction activity nationwide, including the number and value of active projects, the number and value of federally owned projects, the date of the most recent survey in each county, and whether the existing wage determinations for each county are union-prevailing, nonunion-prevailing, or a combination of both. Labor uses the Uniform Survey Planning Procedure to weigh the need for surveys by area and construction type. Once Labor designates an area and construction type (i.e., building, residential, highway, or heavy) for a survey, it proposes a survey time frame, or reference period during which the construction projects considered in the survey must be "active." Generally, the preliminary time frame is the preceding 12-month period, the survey start date is approximately 3 months after the survey is assigned, and the survey cutoff date is 4 to 6 months from the start date, depending on the size of the survey. However, the survey time frame, start date, and cutoff date may be shortened or lengthened based on individual circumstances of the survey. Once these parameters are established, Labor enters the survey information into ASDS. To identify projects that meet the established survey criteria (the designated area, construction type, and survey time frame), Labor uses F.W. Dodge data produced in reports known as Dodge Reports.[Footnote 77] Labor supplements these data with information provided by contractors listed in the Dodge Reports, by industry associations, and from regional office files to find additional relevant construction projects. Analysts at CIRPC screen the data to ensure projects selected meet the criteria before the survey begins. Projects must be of the correct construction type, be in the correct geographic area, fall within the survey time frame, and have a value of at least $2,000. CIRPC also checks for duplicate project information to minimize contacts to a contractor working on multiple projects that meet survey criteria. Stage 2: Conducting Surveys: Labor notifies contractors and interested parties--including contractor associations, unions, government agencies, and Members of Congress--of upcoming surveys by posting survey information on its Web site, sending letters, and conducting pre-survey briefings. Contractor and interested party records are sent to the U.S. Census Bureau, which distributes the notification letters encouraging participation in the survey. Labor's regional offices arrange pre-survey briefings with interested parties prior to or at the start of a survey to clarify survey procedures and provide information on how to complete and submit wage survey forms, known as WD-10s. Data requested on the WD-10 form include a description of the project and its location; the contractor's name and address; the project value and start and end dates; the wage rate and fringe benefits paid to each worker on the project; and the number of workers employed in each job classification during the week of peak activity for that classification. The peak week for each job classification is the week when the most workers were employed in that particular classification. For an example of how Labor collects peak week data on a WD-10, see appendix II. The Census Bureau conducts four mailings throughout a survey. The first mailing includes letters and WD-10 wage survey forms to general contractors and interested parties. (For examples of survey announcement letters sent to contractors and interested parties, see appendix IV.) General contractors listed on the Dodge Reports receive WD-10 forms with project names identified through the Dodge Reports, as well as additional blank forms for other projects. General contractors not listed on the Dodge Reports and interested parties receive a limited number of blank WD-10 forms, but additional forms are available upon request. In addition, all general contractors receive forms to provide information on subcontractors who worked on projects being surveyed. Members of Congress receive one blank WD-10 form and are not contacted again unless a survey is extended. The second mailing is only to general contractors who do not respond to the first mailing and includes the WD-10 forms with project names from the Dodge Report and subcontractor list forms provided in the first mailing. The third mailing is to all reported subcontractors and newly reported general contractors and includes WD-10 forms with project names and blank WD-10 forms. The fourth and final mailing is to all subcontractors who do not respond and newly reported subcontractors and only includes WD-10 forms. Survey respondents may submit paper WD-10 forms or complete forms electronically on Labor's Web site. Census scans returned paper WD-10 forms into Labor's ASDS. WD-10 forms submitted electronically are loaded directly into ASDS. Any additional information submitted must be entered into ASDS manually. CIRPC reviews the completed WD-10s, matches submitted information with the associated project, and forwards the WD-10s to Labor regional offices. Stage 3: Clarifying and Analyzing Respondents' Data: Labor's wage analysts begin to review and analyze the data as they receive the completed WD-10s. Data Review and Clarification: Wage analysts' first step in the review process is to determine whether the project reported on the WD-10 form is within the scope of the survey, or "usable." Since the WD-10 forms may provide more information about a project than the Dodge Report, wage analysts review the data to determine whether the project meets the four basic survey criteria (correct construction type, geographic area, time frame, and project value). If a project does not meet the four criteria, it is determined unusable and any associated WD-10 forms are excluded from the survey. Once Labor has determined a project and WD-10 form are usable, wage analysts call contractors to clarify any information that is unclear or incomplete. Wage analysts record information about the clarification call in ASDS, including the date and name of the person contacted and any information that resulted in changes to the WD-10 form. Wage analysts review each section of the WD-10 forms and clarify the information, as necessary. Specifically, the analysts verify contractor and subcontractor information; project name, description, and location; whether the project received federal or state funding; start and end dates and value of the project; type of construction (i.e., building, residential, highway, or heavy); employee job classifications; the peak week ending date; the number of employees reported; the basic hourly rate; fringe benefits rates; and whether the wages were paid under a CBA, among other data. In addition to contractors, interested parties may also submit WD-10 forms for a project. However, Labor clarifies submitted data with the relevant contractor, regardless of the source, and excludes information provided by an interested party if it duplicates data provided by the contractor unless data are submitted on specific job classifications that were not included by the contractor. Labor also verifies rates paid under a CBA, or union rates, to ensure they are accurately reported. Similarly, because of variations in industry practices across the country, known as "area practice," wage analysts may call contractors to clarify the type of work employees in certain job classifications are actually performing. This is necessary because, for a given prevailing wage, the scope of work covered by the job classification must reflect the actual prevailing area practice. An area practice issue exists when the same work is performed by employees in more than one classification in a given location. For example, a worker under the general electrician classification may perform tasks in addition to general electrical work, such as alarm installation and low voltage wiring. If there is another specialty classification installing alarms in the same location, it may indicate an area practice issue. In some geographic areas, particular work may be performed frequently and widely enough by a specialty classification such that the traditional practice by the general classification may be replaced by the practice of the specialty classification. Data Verification: Labor conducts several processes to verify data submitted in a survey. For data submitted by interested parties and contractors, Labor's regional offices verify a random sample of data. To verify reported data, regional offices contact selected contractors and third parties to request payroll documentation, though data provided without documentation may still be used. In addition to remote verification of randomly selected contractors, on-site verification of a weighted sample of contractors is conducted. The on-site verification selection is designed to include those contractors with the biggest impact on the prevailing wage rate for each job classification.[Footnote 78] Once the weighted sample of contractors has been selected, an independent auditing firm contracted by Labor arranges an appointment with each contractor to meet and review supporting records. The auditing firm prepares and submits a report documenting the differences between the submitted and verified information, including differences in project and wage data. Wage analysts in Labor's regional offices update information in ASDS that may have changed as a result of these verification processes. Additional Data Sources: In addition to wage data collected on WD-10 forms, Labor uses certified payroll data from projects that receive federal funding and meet survey criteria. For surveys of highway and heavy construction projects, Labor always uses certified payroll data, while it is only included in building and residential surveys if the submitted WD-10 forms do not provide enough information to make a wage determination. In addition, for highway surveys only, Labor sometimes adopts rates published by state departments of transportation if a state has conducted its own prevailing wage survey and data collected separately by Labor support the prevailing wage rates established by the state. [Footnote 79] Labor also updates union-prevailing wage rates when unions submit updated CBAs to Labor headquarters. Prevailing Wage Rate Calculation: Once all verified and corrected data have been entered into ASDS, Labor calculates the prevailing wage rate for each job classification in a survey. If a majority of workers (more than 50 percent) in a job classification are paid the same rate, that rate is determined the prevailing wage.[Footnote 80] If the same rate is not paid to a majority (over 50 percent) of workers in a job classification, the prevailing wage is the average wage rate weighted by the number of employees for which that rate was reported. Prevailing fringe benefits are determined only if a majority of the workers in a job classification receive fringe benefits. Once that condition is met, the prevailing fringe benefit is calculated for each job classification similarly to the way the prevailing wage rate is calculated. The prevailing rates resulting from the calculations will be either "union-prevailing"--if a majority of workers is paid under a CBA--or "nonunion-prevailing" rates. A prevailing wage rate for a job classification is only issued if there are sufficient data to make a determination. For data to be sufficient, Labor must receive wage information on at least three employees from at least two contractors for that job classification. If Labor receives sufficient data based on information collected at the county level for a job classification, a prevailing wage rate is determined using data from a single county. If data are insufficient at the county level, Labor includes data from federal projects in that county. If data are still insufficient, Labor includes data from contiguous counties, combined in "groups" or "supergroups" of counties, until data are sufficient to make a prevailing wage determination. Expansion to include other counties, if necessary, may continue until data from all counties in the state are combined. However, Labor's regulations require wage data from projects in metropolitan and rural counties be separated when determining prevailing wages.[Footnote 81] For metropolitan counties, data are combined with data from one or more counties within the metropolitan statistical area, while data from rural counties are combined with data from other rural counties. Once the prevailing wage rates have been calculated, the regional offices transmit survey results to headquarters for final review. Stage 4: Issuing the Wage Determinations: Labor headquarters issues wage determinations after reviewing recommended wage rates submitted by the regional offices. The prevailing wage rates are transmitted electronically to the WDGS for publication online at [hyperlink, http://www.wdol.gov], where they are publicly available. Labor sometimes modifies wage determinations to keep them current or correct errors. Generally, modifications affect a limited number of job classifications within a wage determination. If a prevailing wage rate is not provided for a specific job classification in a wage determination, a contractor may request a rate for that classification, known as a conformance, through the contracting agency overseeing the specific project.[Footnote 82] The rate determined in the conformance process only applies to workers in that classification for the contract in question. Stage 5: Appeals Process: Any interested party may request reconsideration and review of Labor's wage determinations.[Footnote 83] The regional offices accept initial inquiries after a wage determination has been issued. Any interested party may request reconsideration from headquarters in writing and include any relevant information, such as wage payment data or project descriptions, to assist with the review. Labor's regulations state that the Wage and Hour Division Administrator will generally respond within 30 days of receipt of the request.[Footnote 84] If the interested party's request for reconsideration is denied, the interested party may file an appeal with Labor's Administrative Review Board,[Footnote 85] which consists of three members appointed by the Secretary of Labor. All decisions by the Administrative Review Board are final. Any new wage determination resulting from such an appeal must be issued prior to the award of the contract in question, or before the start of construction if there is no award. [End of section] Appendix IV: Survey Announcement Letters Sent to Contractors and Interested Parties by the Department of Labor: Survey announcement letter to contractors identified through Dodge Reports as having active construction projects that meet survey criteria: Davis-Bacon Wage Survey: U.S. Department Of Labor: Employment Standards Administration: Wage & Hour Division: (Date of Letter) (Name of Company) (Address) (City) (State) (Zip Code) Dear Sir or Madam: U.S. Department Of Labor is conducting a wage survey of (Building, Heavy, Highway, and Residential) projects active and ongoing between (Begin Date) through (End Date) in the State of (State). This information will be used to establish prevailing wage rates as required under the Davis-Bacon and related Acts. We understand that your firm is or recently has been engaged in the construction of the project(s) identified on the enclosed form(s). Please complete the WD-10 form for each of the identified projects. If your company has worked on other projects meeting forms for this purpose and a the criteria of this survey, we would appreciate that information also. Blank self-addressed postage paid envelope are included with this letter. Completed WD-10s must be paid by your firm on this project(s) assisted construction projects postmarked by (Return Date) to be included in the survey. The wages being affect the prevailing wage scale established for future federally funded or in this State. In addition to the wages paid any subcontractors you employed A form for this purpose is enclosed. to your employees, it is also important that we receive a list, as soon as possible, of any subcontractors you employed on these projects so that we may contact them and request their participation. Your subcontractor list in any form can also be faxed directly to this office. Please be assured that all information provided will be kept confidential to the maximum extent possible under existing law. Information regarding the Davis-Bacon wage survey program as well as support for completing forms is available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/index.htm]. Project wage data may also be submitted by using the electronic WD-10 which is available at the web site: hyperlink, http://www.dol.gov/whd/programs/dbra/index.htm] Thank you very much for your cooperation in this survey. If you have any questions or need additional forms, please do not hesitate to contact (Analyst) at the office listed below. Sincerely, (Graphic Signature) (Printed Name) Regional Wage Specialist: Enclosures: U.S. Department of Labor: Wage and Hour Division: (Regional Office Address) (City, State, & Zip Code) Phone: (Region phone number) Fax: (Region fax number) [End of letter] Survey announcement letter to contractors not identified through Dodge Reports, but who request to be notified about upcoming surveys: Davis-Bacon Wage Survey: U.S. Department Of Labor: Employment Standards Administration: Wage & Hour Division: (Date of Letter) (Name of Company) (Address) (City) (State) (Zip Code) Dear Sir or Madam: The U.S. Department of Labor is conducting a wage survey of (Building, Heavy, Highway, and Residential) projects active and ongoing between (Begin Date) through (End Date) in the State of (State). This information will be used to establish prevailing wage rates as required under the Davis-Bacon and related Acts. The wages being paid by your firm on any projects may affect the prevailing wage scale for future federally funded or assisted construction projects in this state. If your firm has been engaged in the construction of projects meeting the criteria of this survey, please complete a WD-10 form for each applicable project. WD-10 forms and a self-addressed postage paid envelope are included with this letter. Completed WD-10s must be postmarked by (Return Date) to be included in the survey. In addition to the wages paid to your employees, it is also important that we receive a list, as soon as possible, of any subcontractors you employed on these projects so that we may contact them and request their participation. A form for this purpose is enclosed. Your subcontractor list in any form can also be faxed directly to this office. Please be assured that all information provided will be kept confidential to the maximum extent possible under existing law. Information regarding the Davis-Bacon wage survey program as well as support for completing forms is available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/index.htm]. Project wage data may also be submitted by using the electronic WD-10 which is available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/wd10/index.htm] Thank you very much for your cooperation in this survey. If you have any questions or need additional forms, please do not hesitate to contact (Analyst) at the office listed below. Sincerely, (Graphic Signature) (Printed Name) Regional Wage Specialist: Enclosures: U.S. Department of Labor: Wage and Hour Division: (Regional Office Address) (City, State, & Zip Code) Phone: (Region phone number) Fax: (Region fax number) [End of letter] Survey announcement letter to congressional members: Davis-Bacon Wage Survey: U.S. Department Of Labor: Employment Standards Administration: Wage & Hour Division: (Date of Letter) The Honorable (Representative/Senator) (Address) (City) (State) (Zip Code) Dear (Representative's/Senator's Name): This is to advise you of a Davis-Bacon wage survey to be conducted within your (State/District). The Davis-Bacon and related Acts require payment of locally prevailing wage rates and fringe benefits to employees of contractors and subcontractors performing work on federally financed or assisted construction projects. As you know, the Wage and Hour Division of the U.S. Department of Labor is responsible for conducting wage surveys to establish the wage rates applicable to such construction projects. This office will be collecting wage payment information on any (Building, Heavy, Highway, and Residential) projects within your (State/District). We will be soliciting information of wages and fringe benefits paid on those projects active and ongoing between (Begin Date) through (End Date). Data must be postmarked by (Return Date) to be included in the survey. The wages being paid construction workers in your (State/District) may affect the prevailing wage determinations issued as a result of this survey. The identity of respondents will be kept confidential to the maximum extent possible under existing law. Information regarding the Davis-Bacon wage survey program as well as support for completing forms is also available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbrafindex.htm]. Project wage data may also be submitted by using the electronic WD-10 which is available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/wd10/index.htm] If you have any inquiries from your constituents regarding this survey, we would appreciate your urging their participation. Any questions concerning this survey or relevant information on construction wage rates should be directed to (Wage Specialist) at the office listed below. Sincerely, (Graphic Signature) (Printed Name) Regional Wage Specialist: Enclosures: U.S. Department of Labor: Wage and Hour Division: (Regional Office Address) (City, State, & Zip Code) Phone: (Region phone number) Fax: (Region fax number) [End of letter] Survey announcement letter to interested parties: Davis-Bacon Wage Survey: U.S. Department Of Labor: Employment Standards Administration: Wage & Hour Division: (Date of Letter) (Name of Company) (Address) (City) (State) (Zip Code) Dear Sir or Madam: This is to advise you of a Davis-Bacon wage survey to be conducted within your jurisdiction. The Davis-Bacon and related Acts require payment of locally prevailing wage rates and fringe benefits to employees of contractors and subcontractors performing work on federally financed or assisted construction projects. The Wage and Hour Division of the U.S. Department of Labor is responsible for conducting wage surveys to establish the wage rates applicable to such construction projects. This office will be collecting wage payment information on any (Building, Heavy, Highway, and Residential) projects in the State of (State). We will be soliciting information of wages and fringe benefits paid on those projects active and ongoing between (Begin Date) through (End Date). Data must be postmarked by (Return Date) to be included in the survey. The wages being paid construction workers in this area may affect the prevailing wage determinations issued as a result of this survey. The identity of respondents will be kept confidential to the maximum extent possible under existing law. Information regarding the Davis- Bacon wage survey program as well as support for completing forms is also available at the web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/index.htm]. Project wage data may also be submitted by using the electronic WD-10 which is available at the Web site: [hyperlink, http://www.dol.gov/whd/programs/dbra/wd10/index.htm] Please promptly notify interested parties in your organization and request their participation regarding this survey. We would appreciate your urging their cooperation. Any questions concerning this survey or relevant information on construction wage rates should be directed to (Analyst) at the office listed below: Sincerely, (Graphic Signature) (Printed Name) Regional Wage Specialist: Enclosures: U.S. Department of Labor: Wage and Hour Division: (Regional Office Address) (City, State, & Zip Code) Phone: (Region phone number) Fax: (Region fax number) [End of letter] [End of section] Appendix V: Example of a Florida Wage Determination Published by the Department of Labor: General Decision Number: FL100104: 10/08/2010: FL104: Superseded General Decision Number: F120080104: State: Florida: Construction Type: Building: County: Alachua County in Florida: Building Construction Projects (does not include single family homes or apartments up to and including 4 stories). Modification Number: 0; Publication Date: 03/12/2010. Modification Number: 1; Publication Date: 03/19/2010. Modification Number: 2; Publication Date: 03/26/2010. Modification Number: 3; Publication Date: 05/14/2010. Modification Number: 4; Publication Date: 07/23/2010. Modification Number: 5; Publication Date: 08/06/2010. Modification Number: 6; Publication Date: 10/08/2010. ELEC1205-004: 05/01/2010: Electrician: Rates: $23.48; Fringes: $6.46. * ENGT0925-001: 07/01/2010: Operator: Crane: Crawler Cranes; Truck Cranes; Pile Driver Cranes; Rough Terrain Cranes; and Any Crane not otherwise described below; Rates: $27.91; Fringes: $10.59. Operator: Crane: Hydraulic Cranes Rated 100 Tons or Above but Less Than 250 Tons; and Lattice Boom Cranes Less Than 150 Tons if not described below: Rates: $28.91; Fringes: $10.51. Operator: Crane: Lattice Boom Cranes Rated at 150 Tons or Above; Friction Cranes of Any Size; Mobile Tower Cranes or Luffing Boom Cranes of Any Size; Electric Tower Cranes; Hydraulic Cranes Rated at 250 Tons or Above; and Any Crane Equipped with 300 Foot or More of Any Boom Combination: Rates: $29.91; Fringes: $10.59. Operator: Mechanic Rates: $27.91; Fringes: $10.59. OPERATOR: Oiler Rates: $21.38; Fringes: $10.59. Operator: Boom Truck Rates: $27.91; Fringes: $10.59. Operator: Concrete Pump Rates: $23.41; Fringes: $10.59. IR0N0597-003: 08/01/2009: Ironworker, Ornamental And Reinforcing: Rates: $21.56; Fringes: $7.62. PAIN0088-001: 07/01/2008: Painter: Brush, Roller, Spray and Steel Only: Rates: $16.00; Fringes: $6.85. * PLUMO234-001: 09/01/2010: Plumber/pipefitter: Rates: $26.64; Fringes: $11.54. SHEE0435-005: 07/01/2010: Sheetmetal Worker (Including HVAC Duct Installation): Rates: $22.45; Fringes: $12.38. A: Holiday: 3% of the employee's regular rate of pay times the number of hours worked (excluding fringe benefit contributions), with the first effective holiday beginning Memorial Day, 2008. * SUFL2009-031: 05/22/2009: Bricklayer: Rates: $18.93; Fringes: $0.00. Carpenter, Includes Metal Stud Installation: Rates: $13.47; Fringes: $2.28. Cement Mason/concrete Finisher: Rates: $17.69; Fringes: $1.83. Insulator - Pipe & P1pewrapper: Rates: $13.13; Fringes: $3.03. Ironworker, Structural: Rates: $15.50; Fringes: $0.00. Laborer: Asphalt Shoveler: Rates: $7.88; Fringes: $0.00. Laborer: Common Or General: Rates: $9.45; Fringes: $0.50. Laborer: Concrete Saw: Rates: $12.63; Fringes: $0.00. Laborer: Mason Tender - Brick: Rates: $10.75; Fringes: $0.00. Laborer: Mason Tender - Cement/concrete: Rates: $12.66; Fringes: $1.90. Laborer: Pipelayer: Rates: $8.00; Fringes: $0.00. Laborer: Roof Tearoff: Rates: $8.44; Fringes: $0.00. Laborer: Landscape And Irrigation: Rates: $10.37; Fringes: $0.68. Operator: Asphalt Spreader: Rates: $11.46; Fringes: $0.00. Operator: Backhoe/excavator: Rates: $12.42; Fringes: $0.50. Operator: Bulldozer: Rates: $15.01; Fringes: $0.00. Operator: Distributor: Rates: $13.50; Fringes: $0.00. Operator: Forklift: Rates: $13.50; Fringes: $0.00. Operator: Grader/blade: Rates: $13.73; Fringes: $0.00. Operator: Loader: Rates: $12.20; Fringes: $0.00. Operator: Paver: Rates: $11.20; Fringes: $0.00. Operator: Roller: Rates: $10.59; Fringes: $0.00. Operator: Screed: Rates: $10.77; Fringes: $0.00. Operator: Tractor: Rates: $9.91; Fringes: $0.00. Operator: Trencher: Rates: $11.75; Fringes: $0.00. Roofer (Metal Roofs Only): Rates: $14.26; Fringes: $0.59. Roofer, Including Built Up, Hot Tar, Modified Bitumen, Shake & Shingle, Single Ply and Slate & Tile (Excluding Metal Roof); $13.06 0.00 Tile Setter: Rates: $14.21; Fringes: $1.74. Truck Driver: Dump Truck; Rates: $10.00; Fringes: $0.00. Truck Driver: Lowboy Truck; Rates: $12.16; Fringes: $0.00. Welders - Receive rate prescribed for craft performing operation to which welding is incidental. Unlisted classifications needed for work not included within the scope of the classifications listed may be added after award only as provided in the labor standards contract clauses (29 CFR 5.5(a)(1)(ii)). In the listing above, the "SU" designation means that rates listed--.ender the identifier do not reflect collectively bargained wage and fringe benefit rates. Other designations indicate unions whose rates have been determined to be prevailing. Wage Determination Appeals Process: 1) Has there been an initial decision in the matter? This can be: * an existing published wage determination; * a survey underlying a wage determination; * a Wage and Hour Division letter setting forth a position on a wage determination matter; * a conformance (additional classification and rate) ruling. On survey related matters, initial contact, including requests for summaries of surveys, should be with the Wage and Hour Regional Office for the area in which the survey was conducted because those Regional Offices have responsibility for the Davis-Bacon survey program. If the response from this initial contact is not satisfactory, then the process described in 2) and 3) should be followed. With regard to any other matter not yet ripe for the formal process described here, initial contact should be with the Branch of Construction Wage Determinations. Write to: Branch of Construction Wage Determinations: Wage and Hour Division: U.S. Department of Labor: 200 Constitution Avenue, N.W. Washington, DC 20210: 2) If the answer to the question in 1) is yes, then an interested party (those affected by the action) can request review and reconsideration from the Wage and Hour Administrator (See 29 CFR Part 1.8 and 29 CFR Part 7). Write to: Wage and Hour Administrator: U.S. Department of Labor: 200 Constitution Avenue, N.W. Washington, DC 20210: The request should be accompanied by a full statement of the interested party's position and by any information (wage payment data, project description, area practice material, etc.) that the requester considers relevant to the issue. 3) If the decision of the Administrator is not favorable, an interested party may appeal directly to the Administrative Review Board (formerly the Wage Appeals Board). Write to: Administrative Review Board: U.S. Department of Labor: 200 Constitution Avenue, N.W. Washington, DC 20210: 4) All decisions by the Administrative Review Board are final. End Of General Decision: [End of section] Appendix VI: Comments from the Department of Labor: U.S. Department of Labor: Wage and Hour Division: Washington, D.C. 20210: March 7, 2011: Andrew Sherrill: Director: Education, Workforce, and Income Security Issues: U. S. Government Accountability Office: Washington, D.C. 20548: Dear Mr. Sherrill: Thank you for the opportunity to comment on the Government Accountability Office (GAO) draft report entitled "Methodological Changes Needed to Improve Wage Survey." Many of the concerns discussed in this report are in the process of being addressed by the Department of Labor's (DOL) Wage and Hour Division (WHD) and as noted in this response it is too early to fully assess the effectiveness of WHD's improvements. This letter contains information the GAO may find useful as it considers its final report. Since 2005 the WHD has been working to address recommendations for the Davis-Bacon Act (DBA) wage survey and wage determination process from reviews conducted by the Department of Labor Office of the Inspector General and McGraw Hill Construction Analytics involving improvements in the IT system, timeliness and accuracy of process, personnel, and performance measures. As WHD staff acknowledged to the GAO auditor, these improvements are ongoing. A brief discussion of these efforts follows. Improvements in the IT System: WHD adopted a systematic approach to implement improvements in the wage determinations IT system. From 2005 through January 2011, 29 major releases and updates were made to WHD's Automated Survey Data System (ASDS) and 17 major releases and updates were made to the Wage Determination Generation System (WDGS). The changes were designed to increase the speed of processing so that surveys could be completed and published in a more timely manner. In 2007, a "bridge" connecting both IT systems became operational, which resulted in improvements to performance measurements and reports. For example, preparation of documents for on-site verification was reduced from one month to one day and area practice resolution was reduced from weeks to one day. IT development and subsequent system improvements are still ongoing to further increase the accuracy and timeliness of DBA wage surveys and wage determinations. In addition, reports to assess the performance of IT development and subsequent system improvements have been developed and implemented, and will continue to be reviewed for further improvement. Process (Timeliness and Accuracy): WHD reviewed its survey processes in those key areas in which there was substantial time expenditure. As a result, regional WHD analysts are now performing analysis and clarification of data within two weeks of the receipt of such data; contractor, third party, and on-site verification is being performed within an average of six to eight months from survey cut-off date; and responses from all parties to verification requests have substantially increased. All 22 surveys considered backlogged were completed and published in either FY 2009 or FY 2010. The new survey process for building and heavy construction began in 2009. Of the nine statewide building and heavy construction surveys started in 2009 with data collection cut off dates from December 31, 2009 to February 28, 2010, four (Montana, Wyoming, New Hampshire and Vermont) have been or are being published at this time; three (North Carolina, South Carolina, and Maine) will be published within the next two months; and two (West Virginia and Nebraska) are in on-site verification resulting in an average of 24 months from start of survey to publication of wage determination. As a result of increased relationship building with state Department of Transportation (DOT) offices, 33 states now work with WHD to issue and maintain current prevailing highway wage rates. Upon publication of upcoming surveys, 44 of the 50 highway surveys will have been completed within three years. A survey plan is being developed with a schedule of publishing rates for 17 DOL surveyed states each year so that highway construction wage rates are no older than three years. Improvements have also been made to address difficulty in residential surveys. These surveys will now be conducted separately, allowing for use of strategies to address response and participation rates. WHD began its revised residential construction program in 2010 with a statewide survey of Missouri. Residential surveys of Maine, Vermont, New Hampshire, Maryland, Virginia, North Carolina, South Carolina, Nevada, Washington, and Oregon will follow in 2011 and 2012. Personnel: WHD has increased both its survey staff capacity to provide support for the increase in the number of surveys and the reduced timeframes in which surveys are to be conducted. Training is being emphasized. In 2006 WED instituted a training program for survey staff and an updated manual of operations is pending approval. Improved communication is also a focal point. Yearly planning meetings, monthly conference calls, regular calls, and on-site visits are key components of the communication strategy. Performance Measures: WHD has instituted performance measures for the survey program which address the timeliness of the DBA wage survey and wage determination program. These measures include wage rates, the period of time from completion of the survey to publication, and the time required to conduct surveys. The GAO Draft Report indicates that start dates were being entered into the system differently by regions and, therefore, the ability to accurately measure the survey timeliness was affected accordingly. Any differences in reporting by the regions have been accounted for. In the planned April 2011 release, ASDS will automatically populate these fields when the region enters data so there will be uniformity in reporting. This report along with the analysts' time reports will allow WHD to monitor those processes in which large amounts of time are being spent and allocate resources accordingly. Furthermore, a number of initiatives have resulted in more accurate reporting of information, specifically allowing national office personnel to monitor time spent on specific survey activities, and more closely align staff performance standards with the agency's program performance goals and measures. Responses to Additional Issues Raised and Recommendations: Below are responses to additional issues raised in the Draft Report: Pages 13—25 of the Draft Report discuss survey quality. WHD has implemented numerous changes over the last five years and continues to monitor the survey and wage determination program as discussed briefly above. As GAO acknowledges on page 15 of the Draft Report, "it is too early to fully assess the effects of Labor's 2009 changes." However, we think it is important to note that these changes have indicated improved efficiency thus far. For building and heavy construction, prior to 2009, it took WHD several years to complete and publish a survey. New processes instituted in 2009 and 2010 broke down the survey process for these types of construction with the goal of completing building and heavy construction surveys within 19 months. WHD is now, on average, completing surveys within 24 months. Improvements and changes continue to be made so that WHD may eventually meet this goal. The lack of survey response rate on page 23 is being addressed by WHD. Development of response rate calculations began in 2009. The December 2010 IT release and update provided WHD the ability to trace responses for every contractor and interested party. The subsequent April 2011 release will give reporting capability. The breakdown of this information by construction type will be included in a future update to be released in late 2011 or early 2012. Pages 26—37 of the Draft Report address incentives to participate and transparency as key issues for stakeholders. As the Draft Report notes, participation in the survey process is voluntary and there is no penalty for providing inaccurate information. These circumstances may affect incentives to participate. WHD understands that any confusion about the survey form by any stakeholder is undesirable. However, the data errors discussed in the Draft Report typically (if not always) result from errors in the information provided by survey respondents, not from errors by WHO employees. There is no indication in the Draft Report that these error rates had any impact on the accuracy of the wage determinations themselves. WHD agrees that greater transparency will enhance the process and it has already. Page 29 of the Draft Report implies that errors may have occurred because WHD did not pretest a redesigned form. However, this form is not a new form. It is the old WD-10 data placed in a scannable format. The scannable format eases the information collection process for the participant. Regarding outreach concerns, contact is made with unions and contractor associations in every survey. For Florida and New York mentioned in the Draft Report, the contractor associations did not respond to WHD offers of pre-survey briefings. WHD will continue to work with the contractor associations, unions, and other interested parties to ensure that every effort is made to increase participation and solicit the necessary wage information. As surveys are conducted more regularly, WHD anticipates that participating in these surveys will become routine for the stakeholders, thus decreasing confusion and increasing the overall response rates. The Draft Report contains two recommendations for DOL, The responses for which are as follows: Recommendation #1: WHD has previously enlisted McGraw Hill Construction Analytics to assess WHD's process and operations. The recommendations from McGraw Hill have been implemented and are beginning to show results. Given that further changes to the process are currently being implemented or will be implemented in the near future, contracting to a different organization to evaluate the efforts of WHD may be premature, especially in light of cost considerations. Recommendation #2: WHD agrees with the recommendation that the public should have more information to clearly understand the information being requested and the calculations and codes that are used on the wage determinations and is already undertaking steps to address these concerns. The wage determinations are housed on the Website "WDOL." WDOL was a collaboration of the Department of Labor, OMB, National Technical Information Service, the General Services Administration and the Department of Defense. Any changes to the Website must be made in collaboration with these other entities and cannot be made unilaterally by DOL. As noted in this response, the concerns GAO identifies in this Draft Report regarding the wage determination process are being addressed by WHD. Many of the changes to correct the outlined issues are in their infancy and their effectiveness cannot yet be fully assessed. WHD will continue to track the effectiveness of its actions to improve the wage determination process. Thank you for providing us with the opportunity to comment on the Draft Report. If you have any questions, please do not hesitate to contact us. Sincerely, Signed by: Thomas M. Markey: Deputy Administrator Program Operations: [End of section] Appendix VII: GAO Contact and Staff Acknowledgments: GAO Contact: Andrew Sherrill (202) 512-7215 or sherrilla@gao.gov: Staff Acknowledgments: In addition to the contact named above, the following staff made key contributions to this report: Gretta L. Goodwin, Assistant Director and Amy Anderson, analyst-in-charge, managed all aspects of this assignment; and Brenna Guarneros, analyst, made significant contributions to all phases of the work. In addition, John J. Barrett, analyst, made significant contributions to design and data collection; Christopher Zbrozek, intern, assisted in data collection and analysis; Walter Vance, Melinda Cordero, and Carl Barden provided assistance in designing the study and conducting data analysis; Susan Aschoff assisted in message and report development; Mimi Nguyen created the report's graphics; Alexander Galuten provided legal advice; Erin Godtland, Barbara Steel-Lowney, and Yunsian Tai referenced the report; and Roshni Dave, Ronald Fecso, Kim Frankena, Mark Gaffigan, Charles A. Jeszeck, David Marroni, Mary Mohiyuddin, Stuart Ryba, David Wise, and William Woods provided guidance. [End of section] Footnotes: [1] GAO, Davis-Bacon Act: Process Changes Could Raise Confidence That Wage Rates Are Based on Accurate Data, [hyperlink, http://www.gao.gov/products/GAO/HEHS-96-130] (Washington, D.C.: May 31, 1996) and Davis-Bacon Act: Labor Now Verifies Wage Data, but Verification Process Needs Improvement, [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-21] (Washington, D.C.: Jan. 11, 1999). [2] GAO, Davis-Bacon Act: Labor's Actions Have Potential to Improve Wage Determinations, [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-97] (Washington, D.C.: May 28, 1999). [3] Department of Labor, Office of Inspector General, Concerns Persist with the Integrity of Davis-Bacon Prevailing Wage Determinations, 04- 04-003-04-420 (Washington, D.C., Mar. 30, 2004). [4] Pub. L. No. 111-5, 123 Stat. 115. [5] The Congressional Budget Office estimated in early 2009 that the combined spending and tax provisions of the Recovery Act would cost $787 billion from 2009 through 2019. In April 2009, the Congressional Research Service estimated the budget authority for Division A of the act to be more than $300 billion for the same time period. Division A consists primarily of discretionary spending, with some exceptions. [6] In the OMB guidance, "objectivity" focuses on whether the disseminated information is being presented in an accurate, clear, complete, and unbiased manner; "utility" refers to the usefulness of the information to the intended user; and "integrity" refers to the protection of information from unauthorized access or revision, and to ensure that the information is not compromised through corruption or falsification. For the purposes of our report, we are defining "quality" by these OMB guidelines. See Office of Management and Budget, Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated by Federal Agencies (effective date Jan. 3, 2002). For additional OMB guidance on agency surveys see Office of Management and Budget, Standards and Guidelines for Statistical Surveys (September 2006) and Office of Management and Budget, Questions and Answers When Designing Surveys for Information Collections (January 2006). For Labor guidance, see Department of Labor, Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated by the Department of Labor (Oct. 1, 2002). [7] The other two regional offices are the Midwest region (Chicago) and the West region (San Francisco). [8] See appendix I for information on the construction types included in the surveys we analyzed. [9] Pub. L. No. 71-798, 46 Stat. 1494 (1931), as amended; codified at 40 U.S.C. § 3141 et seq. [10] 40 U.S.C. §§ 3142, 3141(2)(B). Laborers and mechanics include those workers whose duties are manual or physical in nature as distinguished from mental or managerial duties. 29 C.F.R. § 5.2(m). [11] Pub. L. No. 111-5, 123 Stat. 115 § 1606 (2009). [12] GAO, Recovery Act: Officials' Views Vary on Impacts of Davis- Bacon Act Prevailing Wage Provision, [hyperlink, http://www.gao.gov/products/GAO-10-421] (Washington, D.C.: Feb. 24, 2010). [13] Office of Management and Budget, Analytical Perspectives, Budget of the United States Government, Fiscal Year 2011, Special Topics, "Table 20-2. Federal Investment Budget Authority and Outlays: Grant and Direct Federal Programs." [14] Interested third parties include contractor associations; labor unions; federal, state, and local agencies; and Members of Congress. [15] Surveys are conducted statewide except in large states, such as Texas and California. [16] [hyperlink, http://www.gao.gov/products/GAO/HEHS-96-130]. [17] Department of Labor, Office of Inspector General, Inaccurate Data Were Frequently Used in Wage Determinations Made Under the Davis-Bacon Act, 04-97-013-04-420 (Washington, D.C., Mar. 10, 1997). [18] BLS is part of the Department of Labor. Its mission is to collect, analyze, and disseminate essential economic information to support public and private decision making. [19] [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-21]. [20] [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-97]. [21] "Peak week" refers to the work week in which the contractor employed the largest number of workers in a particular job classification for a specific construction project. [22] The Davis-Bacon wage survey process described here is based on Labor regulations (29 C.F.R. § 1.1 et seq.), procedures manuals and documents, and statements by officials. GAO did not verify whether all procedures were followed in all cases. [23] McGraw-Hill Construction publishes the Dodge Reports, which provide information on the project type, project value, and contractors. [24] The five Labor regional offices that process Davis-Bacon wage surveys are Northeast (Philadelphia), Southeast (Atlanta), Midwest (Chicago), Southwest (Dallas), and West (San Francisco). [25] Contractors or interested parties who submitted wage data but have declined to provide payroll documentation in response to three requests will be kept on a list for 1 year during which time their data will not be used to calculate wage rates. However, because of the length of time between surveys, a senior Labor official said it is unlikely a contractor or interested party would be surveyed frequently enough to be put on the list. [26] 29 C.F.R. § 1.2 (a)(1). [27] Key job classifications across all four construction types include bricklayer, boilermaker, carpenter, cement mason, electrician, heat and frost insulator/asbestos worker/pipe insulator, iron worker, laborer-common, painter, pipefitter, plumber, power equipment operator, roofer, sheet metal worker, tile setter, and truck driver. [28] 40 U.S.C. § 3142(b). [29] 29 C.F.R. § 1.7(a). [30] 29 C.F.R. § 1.7(b). [31] 29 C.F.R. §§ 1.8, 1.9. [32] A Labor official said the agency started planning new surveys again in late 2008. [33] Contractors and subcontractors working on federally funded or assisted construction projects are required to submit weekly payrolls, referred to by Labor as "certified payrolls," to an agency in charge at the site of the work. A Labor official said state departments of transportation collect certified payrolls and provide them to Labor upon request. Assessing the accuracy of certified payrolls was not part of our review. [34] Although Labor generally does not conduct on-site verification of certified payroll data, the agency's revised timeline includes 1 day to review data submitted by interested parties and determine whether they require on-site verification. [35] The building and heavy surveys were conducted in the states of Montana, North Carolina, South Carolina, West Virginia, and Wyoming. [36] Labor officials said they used Recovery Act funding to update 11 of the oldest highway surveys under the new survey process. [37] The 14 state surveys were Arkansas, Connecticut, Florida, Georgia, Iowa, Minnesota, New Jersey, New York Rural, Tennessee, Texas Rural, District of Columbia, West Texas Metropolitan, Utah, and Washington. [38] This includes building, heavy, and highway surveys only. In addition, for building and heavy surveys the estimated time period includes data collection, as data collection, clarification, and analysis happen concurrently under Labor's new processes. We were unable to assess time frames for data clarification and analysis for surveys started under the new processes because, according to a senior official, Labor manually records the dates a survey enters and exits a specific processing activity at the end of the survey. [39] While a census survey attempts to collect data from the entire population, a sample survey collects data from a subset or sample of the population. When the sample is selected by a probability sampling method such that each member of the population has a known chance of being selected and that information is used with proper estimation techniques, the results are generalizable to the entire population with a known level of confidence in the precision of the estimates. Further, by reducing the data collection effort, more can be done to assure other aspects of data quality. [40] The state surveys with data 5 or more years old when results were published are Arkansas 2004 (heavy and residential construction types), Florida 2005 (heavy and residential), Maryland 2005 (all construction types), Missouri 2005 (building), New Jersey 2004 (building, heavy, and residential), and Oregon 2004 (building, heavy, and residential). The state surveys with data 7 or more years old when results were published are Connecticut 2002 (building, heavy, and residential), Minnesota 2002 (building, heavy, and residential), and Washington 2002 (building, heavy, and residential-metropolitan). [41] Labor officials said that to avoid issuing a wage determination with a large number of missing wage rates for job classifications, the agency also requires sufficient wage rate data for at least 50 percent of the key job classifications for each construction type in order to publish a wage determination. [42] Davis-Bacon wage rates include all four construction types: building, heavy, highway, and residential. For BLS unionization figures, see Bureau of Labor Statistics, Economic News Release: Union Members-2010, "Table 3. Union affiliation of employed wage and salary workers by occupation and industry" (Jan. 21, 2011). [43] According to Labor documentation, initial steps to update the survey database started in December 2010, but a final implementation date had not yet been determined as of November 18, 2010. [44] 29 C.F.R § 1.7(a). [45] We analyzed wage rates for key job classifications because wage rates for nonkey job classifications can only be issued at the county or group level, but not at the supergroup or state level. [46] Regional office officials said they may combine rates from counties with the exact same wage and fringe benefit data in the final WD-22, which is a wage compilation report. However, the rates being combined may have been calculated at different geographic levels--for example, one county's rates may have been calculated at the group level while another county's rates may have been calculated at the supergroup level. The geographic level at which rates for combined counties were calculated is not reported on the WD-22; therefore, we reported the percentage of these rates separately. [47] 40 U.S.C. § 3142(b). [48] A metropolitan statistical area (MSA) is an area containing a large population center and adjacent communities that have a high degree of integration with that center. Specifically, an MSA is a statistical area associated with at least one urbanized area that has a population of at least 50,000. An MSA comprises the central county or counties containing a densely settled concentration of population plus adjacent outlying counties having a high degree of social and economic integration with the central county or counties as measured through commuting. [49] The Bureau of Economic Analysis is an agency within the Department of Commerce. It collects source data, conducts research and analysis, develops and implements estimation methodologies, and disseminates economic statistics to the public. [50] We interviewed a total of 30 stakeholders; however, one stakeholder did not address this issue. [51] On the wage survey, Labor enters the name of construction projects it has identified through Dodge Reports, but also asks contractors to provide information on other construction projects they may be working on that meet survey criteria. [52] Labor uses federal data in all highway and heavy surveys because of the high percentage of federal projects in both these construction types. [53] We did not ask the representatives from academia about the form because they generally would not be asked to fill out the form as a survey respondent. [54] The four surveys we reviewed--Florida 2005, Maryland 2005, Tennessee 2006, and West Texas Metropolitan 2006--were conducted prior to new survey processes being implemented. No verification reports for surveys conducted under the new processes were available in time for our review. [55] Labor's contractor asks employers the following questions during on-site verification: (1) Did you have difficulty understanding the information requested on the WD-10 form? (2) Do you have any suggestions for the improvement of the WD-10? (3) Were there specific blocks on the form that you found particularly confusing or difficult to complete? (4) Are you aware that you can complete the WD-10 form on the computer screen and send it in by Internet? [56] The Web site address is [hyperlink, http://www.wdol.gov]. A search for a wage determination can be defined by state, county, construction type, or wage determination number. [57] 29 C.F.R. § 1.5(b)(1). When requesting the wage rate for a classification not listed on the wage determination, agencies must provide a description of the work and construction type, the county (or other civil subdivision) and state where the proposed project is located, and any pertinent wage payment information that may be available. Conformances are requested after a federal construction project has been bid and awarded and generally take at least 30 days to process. [58] During our review, the wage surveys in Arizona, North Carolina, and West Virginia were ongoing, with Labor officials clarifying and verifying survey data. While we interviewed contractor associations in these states, we did not directly interview contractors so as not to inadvertently affect Labor's survey results. See appendix I for more information. [59] We did not ask the representatives from academia about pre-survey briefings because they would generally not be one of the groups Labor would notify about an upcoming survey. [60] Department of Labor, Office of Inspector General, Concerns Persist with the Integrity of Davis-Bacon Prevailing Wage Determinations, 04-04-003-04-420 (Washington, D.C., Mar. 30, 2004); and Department of Labor, Office of Inspector General, Inaccurate Data Were Frequently Used in Wage Determinations Made Under the Davis-Bacon Act, 04-97-013-04-420 (Washington, D.C., Mar. 10, 1997). [61] GAO, Davis-Bacon Act: Process Changes Could Raise Confidence That Wage Rates Are Based on Accurate Data, [hyperlink, http://www.gao.gov/products/GAO/HEHS-96-130] (Washington, D.C.: May 31, 1996); Davis-Bacon Act: Labor Now Verifies Wage Data, but Verification Process Needs Improvement, [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-21] (Washington, D.C.: Jan. 11, 1999); and Davis-Bacon Act: Labor's Actions Have Potential to Improve Wage Determinations, [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-97] (Washington, D.C.: May 28, 1999). [62] Letter from Assistant Secretary for Employment Standards, U.S. Department of Labor to Subcommittee on Labor, Health and Human Services, and Education, Committee on Appropriations, United States Senate and Subcommittee on Labor, Health and Human Services, Education and Related Agencies , Committee on Appropriations, U.S. House of Representatives, Jan. 17, 2001; letter from Deputy Secretary, U.S. Department of Labor to Director, Office of Management and Budget, Aug. 8, 2002; and letter from Assistant Secretary for Employment Standards, U.S. Department of Labor to Chairman, Committee on Appropriations, United States Senate, May 8, 2006. [63] We selected a cutoff date of September 1, 2010, to allow time for data review and analysis. [64] See Office of Management and Budget, Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated by Federal Agencies (effective Jan. 3, 2002); Office of Management and Budget, Standards and Guidelines for Statistical Surveys (September 2006); Office of Management and Budget, Questions and Answers When Designing Surveys for Information Collections (January 2006); and U.S. Department of Labor, Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated by the Department of Labor (Oct. 1, 2002). [65] The WD-22 provides, for each job classification, information on the final prevailing wage and fringe benefit rates, the total number of workers reported, the method of rate calculation--for example, whether the rate was based on a majority or an average--and whether federal data were used in the rate calculation. [66] When data from a county are insufficient to issue a wage rate for a job classification, a group of counties is created by combining a rural county's data with data from one or more contiguous rural counties. A metropolitan county's data are combined with data from counties within the metropolitan statistical area (MSA). When data from a group are still not sufficient to issue a wage rate, a supergroup wage rate is created by combining a rural county's data with data from additional contiguous rural counties, while a metropolitan county's data are combined with county data from other MSAs or the consolidated MSA counties. Finally, if supergroup wage data are not sufficient to issue a wage rate, a statewide rate is created by combining the data for all rural counties or all metropolitan counties in the state. Counties are combined based on whether they are metropolitan or rural, and cannot be mixed. [67] Labor's fiscal year 2010 performance plan goal was for 90 percent of wage rates for building, heavy, and highway construction types to be no more than 3 years old. [68] The two remaining Wage and Hour Division regional offices that process Davis-Bacon wage survey forms are the Midwest region (Chicago) and the West region (San Francisco). [69] Bureau of Labor Statistics, Economic News Release: Union Members - 2010, "Table 5: Union affiliation of employed wage and salary workers by state" (Jan. 21, 2011). According to BLS, the percentage of all employed workers represented by unions is 8.1 percent in Arizona, 26 percent in New York, 4.9 percent in North Carolina, and 16.5 percent in West Virginia. [70] Survey data collection for the Florida 2005 survey and the New York Rural 2006 survey was completed in 2006. According to BLS's 2010 report, the percentage of all New York workers represented by unions is 26 percent, the highest in the nation, while the percentage of all Florida workers represented by unions is 6.9 percent. [71] 40 U.S.C. §§ 3141(2)(B), 3142. [72] The process described here is based on Labor regulations, procedures manuals and documents, and statements by officials. GAO did not verify whether all procedures were followed in all cases. [73] Heavy construction is a catch-all grouping that includes projects not properly classified under the other three types of construction; for example, dredging and sewer projects. [74] 29 C.F.R. § 1.7(a),(b). [75] A wage determination is the listing of wage and fringe benefit rates for each job classification of workers that the Wage and Hour Division Administrator has determined to be prevailing in a given area for a type of construction. [76] Labor's fiscal year 2010 performance goal was for 90 percent of wage rates for building, heavy, and highway construction types to be no more than 3 years old. [77] McGraw-Hill Construction publishes the Dodge Reports. [78] Contractors with the biggest impact include those whose data were used for a job classification with enough data to issue a prevailing wage rate. Of those, contractors with the most employees are selected for on-site verification. [79] Labor also has procedures for surveys of specialized construction, such as dam and dredging projects, as well as American Indian reservation construction. [80] 29 C.F.R. § 1.2(a)(1). [81] 29 C.F.R. § 1.7(b). [82] See 29 C.F.R. § 1.5(b)(1). [83] 29 C.F.R. §§ 1.8, 1.9. [84] 29 C.F.R. § 1.8. [85] 29 C.F.R. § 1.9. [End of section] GAO's Mission: The Government Accountability Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, go to [hyperlink, http://www.gao.gov] and select "E-mail Updates." Order by Phone: The price of each GAO publication reflects GAO’s actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO’s Web site, [hyperlink, http://www.gao.gov/ordering.htm]. Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537. Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information. To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: E-mail: fraudnet@gao.gov: Automated answering system: (800) 424-5454 or (202) 512-7470: Congressional Relations: Ralph Dawn, Managing Director, dawnr@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, D.C. 20548: Public Affairs: Chuck Young, Managing Director, youngc1@gao.gov: (202) 512-4800: U.S. Government Accountability Office: 441 G Street NW, Room 7149: Washington, D.C. 20548: