Congress Should Fix the Flawed Wage Determination Process Before Expanding the Davis-Bacon Act

Report Jobs and Labor

Congress Should Fix the Flawed Wage Determination Process Before Expanding the Davis-Bacon Act

December 13, 2007 5 min read Download Report
James Sherk
James Sherk
Research Fellow, Labor Economics
As research fellow in labor economics at The Heritage Foundation, James Sherk researched ways to promote competition and mobility.

Eight sections of the Energy Independence and Security Act of 2007 (EISA, H.R. 6) would greatly extend the reach of the Davis-Bacon Act. Under Davis-Bacon, Congress requires federal construction contractors to pay "prevailing" wages to prevent the government's buying power from forcing down wages of construction workers, but most of these sections do not concern federal construction projects. Additionally, Davis-Bacon wages bear little relation to the wage rates that prevail in the marketplace. Specifically, Davis-Bacon wage determinations are not made on the basis of scientifically valid random samples, and investigators have uncovered errors in 100 percent of the wage reports they reviewed. Further, the Department of Labor takes several years to process and publish wage determinations. Rather than expand its coverage, Congress should amend the Davis-Bacon Act to require prompt wage determinations to be made using proper statistical techniques.

A Massive Expansion of Davis-Bacon

The Davis-Bacon Act requires contractors on federal construction projects to pay the "prevailing" wage, the typical wage that workers in that area would earn on private construction projects. Eight sections of the Energy Independence and Security Act extend the coverage of the Davis-Bacon Act:

  • § 136: Advanced Technology Vehicles Manufacturing Incentive Program
  • § 471: Energy Sustainability and Efficiency Grants and Loans for Institutions
  • § 491: High Performance Green Buildings Demonstration Project
  • § 545: Energy Efficiency and Conservation Block Grants
  • § 702: Carbon Capture and Sequestration Research, Development, and Demonstration Program
  • § 803: Renewable Energy Construction Grants
  • § 1112: Capital Grants for Class II and Class III Railroads
  • § 1506: New Clean Energy Renewable Energy Bonds

Few of these programs involve federal construction. § 136, for example, provides funding for automobile manufacturers to re-equip their plants to make more efficient vehicles, and § 471 provides loans to institutions to increase their energy efficiency. This bill would thus extend Davis-Bacon coverage from federal construction projects to private projects that Congress has chosen to subsidize.

Government Buying Power Not a Concern

Congress passed the Davis-Bacon Act during the Great Depression, when the federal government was the largest construction contractor in many areas of the country. It had what economists term "monopsony power"--the power, as the biggest buyer of construction services, to force construction workers to work for lower wages. Congress required federal construction projects to pay prevailing wages to prevent the government's purchasing power from driving down wages.

Monopsony power, however, is not a concern in private construction projects that receive some federal funding, such as those contained in EISA. No individual private buyer has the purchasing power of the federal government, so Congress has no economic justification to extend Davis-Bacon coverage to private construction.

Unscientific Surveys

Before Congress extends Davis-Bacon coverage to private construction projects, it should fix the problems that the Department of Labor's inspector general has identified with its implementation. Repeated investigations have found severe flaws in prevailing wage calculations.

The Department of Labor's Wage and Hour (WH) Division conducts surveys of prevailing construction wages. Unlike Bureau of Labor Statistics (BLS) surveys that estimate the unemployment rate or average wages, the Davis-Bacon survey is not a statistically random sample. The survey is self-reported, which means that only those construction firms that take the time to fill it out and send it back influence the survey.[1] This introduces considerable bias into the estimates. Many smaller construction contractors without the staff resources to devote to government paperwork and contractors who do not do business with the federal government do not return the survey.

Surveys must be random in order to provide accurate estimates. Internet polls often show candidates winning 80 percent of the vote--even when those candidates go on to lose the election--because the respondents to those polls are self-selected. Any survey conducted without random sampling reports severely biased results.

Widespread Data Errors

In addition to sampling bias, government investigations have found widespread errors in the wage reports submitted to the Department of Labor. The Government Accountability Office (GAO) found repeated errors in the mid- and late 1990s that led the Wage and Hour Division to overhaul its methodology. In a follow-up report, investigators from the Office of the Inspector General found "one or more errors existed in 100 percent of the wage reports they reviewed. Error rates were high even after WH's prolonged efforts to edit and clarify and complete the data."[2] Wage rate misreporting, employee miscounts, and incorrect job classifications were common.[3] The government does not use accurate data to calculate prevailing wages.

Long Survey Processing Delays

Delays also plague prevailing wage determinations. The Department of Labor takes an average of 2.3 years to issue a prevailing wage determination after the survey period ends. WH spends four-fifths of this time on data correction and analysis, not data collection.[4] Inflation and changing market conditions mean that even perfectly accurate wage determinations do not reflect market wages. Construction contractors must pay "prevailing" wages that are several years out of date.

Highly Inaccurate Results

Self-reported surveys that are two years out of date with 100 percent error rates do not reflect prevailing wages. Table 1 compares Davis-Bacon wages with scientific surveys of median market wages conducted by the Bureau of Labor Statistics for several U.S. cities and construction occupations.

In almost every case, the prevailing wages do not resemble the actual market wages. Davis-Bacon wages vary from 38 percent below market wages for electricians in the Tampa Bay area to 73 percent above market wages for plumbers in San Francisco. Among these cities, Davis-Bacon wages vary an average of 33 percent from market wages.[5] The Davis-Bacon Act, as currently enforced, causes the government to drive down construction wages in some cities and requires taxpayers to overpay for construction projects in others.

Use Scientifically Valid Surveys

Congress should not extend the Davis-Bacon Act while such severe flaws plague the program. Instead, Congress should require the Department of Labor to base prevailing wage determinations on timely, accurate, and scientifically valid surveys.

BLS conducts many such surveys to calculate unemployment, the inflation rate, and other important economic statistics. Two BLS surveys--the National Compensation Survey and the Occupational Employment Statistics (OES) survey--calculate detailed wage information by occupation throughout the United States and could be used to calculate Davis-Bacon prevailing wages. The government already uses OES data to calculate prevailing wages for service occupations under the Service Contract Act.

The Department of Labor does not use BLS surveys to calculate Davis-Bacon prevailing wages, because the Davis-Bacon Act requires prevailing wage calculations for individual cities and counties. The Bureau of Labor Statistics conducts its surveys on the basis of larger metropolitan statistical areas (MSAs) and non-metropolitan statistical areas. Many cities and counties are too small for individual random sampling. MSAs and non-MSAs are economically similar areas that are large enough to generate reliable, statistically valid estimates. Congress should amend the Davis-Bacon Act to require statistically valid random sampling of MSAs, instead of civil subdivisions, so that the Department of Labor can calculate accurate prevailing wages.

Recommendations for Congress

Congress should not extend the Davis-Bacon Act to private construction projects that receive federal subsidies. Private construction has nothing to do with the purpose of prevailing wage laws, which is preventing the federal government's purchasing power from driving down wages. In addition, serious flaws in prevailing wage determinations mean that Davis-Bacon does not serve this purpose. In some cities, Davis-Bacon wages drive down market wages, while in others they cause taxpayers to overpay for construction projects.

Instead Congress should fix the Davis-Bacon prevailing wage determinations by requiring the Department of Labor to calculate prevailing wages using proper statistical techniques and allowing the Department of Labor to calculate prevailing wages using larger geographic areas than civil subdivisions in order to generate statistically valid random samples.

The Bureau of Labor Statistics conducts timely, accurate, and scientifically valid surveys. Using these surveys would significantly improve the accuracy of Davis-Bacon prevailing wages so that they would reflect actual prevailing market wages.

James Sherk is Bradley Fellow in Labor Policy in the Center for Data Analysis at The Heritage Foundation.

[1]U.S. Department of Labor, Office of Inspector General, "Concerns Persist with the Integrity of Davis-Bacon Act Prevailing Wage Determinations," March 30, 2004, pp. 12-13, at

[2]Ibid, p. 1.

[3]Ibid, pp. 11-12.

[4]Ibid, pp. 13-15.

[5]Author's calculations using data from Table 1.


James Sherk
James Sherk

Research Fellow, Labor Economics