Prevailing Wage

THE EFFECTS OF PROJECT LABOR AGREEMENTS IN MASSACHUSETTS by Jonathan Haughton, Ph.D., Darlene C. Chisholm, Ph.D. and Paul Bachman, MSIE* Beacon Hill Institute

Executive Summary

The Beacon Hill Institute has completed a groundbreaking study of the effects of Project Labor Agreements (PLAs) on the costs of school construction projects. In our analysis of 54 school construction projects undertaken in the Greater Boston area since 1995, we find that costs are substantially higher when a school construction project is executed under a PLA. After adjusting the data for inflation, using an index that includes the trend in both construction wages and in materials costs, and after controlling both for the size of projects and for whether they involve new construction or renovations, we find that the presence of a PLA increases project costs by $37.88 per square foot (in 2001 prices) relative to non-PLA projects.

This price differential represents 22.1% of the cost ($171 per square foot) of the average PLA project and amounts to an average of more than $5.0 million out of the $22.7 million cost per school. To illustrate: Malden could have saved $5.3 million on the Beebe School, Wilmington could have saved over $5.7 million on its new middle school, and Milton could have saved almost $6.6 million on its new high school had those projects not been bid under PLAs. Our findings show that the potential savings from not entering a PLA on a school construction project range from $1.9 million for a 50,000-square-foot structure to $9.5 million for a 250,000-square-foot structure.

Massachusetts policymakers and taxpayers need to consider this substantial additional cost in determining whether a PLA agreement is best for school construction in their towns or school districts.

I. Introduction

Project Labor Agreements (PLAs) discourage non-union contractors from bidding on state construction projects by requiring them to conform to union rules and hire through union halls. It is widely believed that construction projects are more expensive when a PLA is in effect. Until now, however, the evidence for this has been largely anecdotal. No one has, to our knowledge, attempted carefully to estimate the degree to which PLAs might increase project costs.

The current study breaks new ground in that it presents clear evidence on the differences in cost per square foot between PLA and non-PLA projects. This measure is based on a thorough study of the cost of school construction projects in the greater Boston area since 1995. We use our measure to estimate the cost savings that would have accrued to three Massachusetts communities had their recent school construction projects not involved PLAs.

II. Historical Background of PLAs

Project Labor Agreements are a form of “pre-hire” collective bargaining agreement between the construction clients (such as towns or school districts) and labor unions pertaining to a specific project, contract or work location, and are unique to the construction industry. The terms of PLAs generally recognize the participating unions as the sole bargaining representatives for the workers covered by the agreements, regardless of current union membership status of these workers. A PLA requires all workers to be hired through the union hall referral system. Non-union workers must join the signatory union of their respective craft and pay dues for the length of the project. The workers’ wages, pension contributions and working hours and the dispute resolution process and other work rules are also prescribed in the agreement. PLAs supersede all other collective bargaining agreements and prohibit strikes, slowdowns and lockouts for the duration of the project.1

Project Labor Agreements in the United States originated in the public works projects of the Great Depression, which included the Grand Coulee Dam in Washington State in 1938 and the Shasta Dam in California in 1940. PLAs have continued to be used for large construction projects since World War II, including the construction of Cape Canaveral in Florida, the current central artery project (the “Big Dig”) in Boston, and even private projects, such as the Alaskan pipeline and Disney World in Florida.

PLAs in the Balance

As PLAs have become more common in publicly financed construction projects, and as the number of non-union construction firms has grown, PLAs have become controversial. Opponents of PLAs argue:

  1. that the agreements raise the cost of undertaking projects, and
  2. that non-union or open-shop contractors are discouraged from bidding on jobs that have PLAs.

Opponents cite the PLA requirements that all employees must be hired in union halls, pay union dues, contribute to union-sponsored retirement plans, and follow union work rules. They argue that the use of a union hiring hall can force the contractors to hire union workers over their own work force. The contractors and their employees are required to pay union wages, dues and contributions into union benefit plans even if covered by their own plans. The work rules restrict the contractors from using their own more flexible operating rules and procedures. These restrictive conditions cause costs to rise for a project that requires a PLA. In passing, it is worth noting that whether or not a PLA is in effect, all contractors must adhere to any “prevailing wage” rules that may be in effect.

Furthermore, open-shop contractors contend that their competitive advantages are nullified by the PLA. The result is that in practice, if not in principle, they are unable to bid competitively on jobs that have a PLA requirement. In turn, the absence of open-shop bidders for PLA projects results in fewer bidders for the project, and with fewer bidders, the lowest bids come in higher than if open-shop contractors had participated. Therefore, the cost of the project will be higher, with fewer bidders attempting to under-bid each other for the contract. Some opponents also argue that requiring a PLA violates state competitive bidding laws that require a free and open bidding process. A number of critics even see PLAs as a form of extortion, with an implicit threat that if a town does not agree to a PLA, then there is more likely to be disruption at the workplace.

Proponents of PLAs claim that the agreements provide for work conditions that are harmonious, and that they guarantee wage costs for the life of the contract. They contend that the provisions that prohibit strikes, slowdowns and lockouts keep the project on time and prevent cost overruns due to delays. Furthermore, the wage stipulations allow firms accurately to estimate labor costs for the life of the project and thus have more accurate bids. Also, the union rules allow for a safer work environment, thereby reducing accidents and thus lowering the number of workman’s compensation claims. Workers’ union certifications ensure the quality of the work and save money by avoiding costly mistakes. Or so it is argued.

The controversy over the use of PLAs in public construction projects has become more intense since the late 1980’s. Open-shop (non-union) construction firms and industry organizations have challenged PLAs in the courts. As discussed below, the executive and legislative branches at the federal, state and local levels of government have at times taken positions in favor of the use of PLAs.

PLAs at the Federal Level

The executive branch of the federal government has been involved in the PLA debate for over a decade. The administration of George H. W. Bush issued an Executive Order in 1992 forbidding the use of PLAs on federally funded projects.2 The Clinton Administration rescinded that order in February 1993 and attempted to go further in 1997, when it planned to issue an executive order requiring all federal agencies to use PLAs on their construction projects. However, due to extensive lobbying, the President instead issued a memorandum encouraging the use of PLAs on contracts over $5 million for construction projects, including renovation and repair work, for federally owned facilities.3 President George W. Bush canceled the Clinton order on February 17, 2001 by issuing an executive order prohibiting PLAs on federally funded and assisted construction projects.4

PLAs in Massachusetts

In Massachusetts, PLAs appeared on the legislative agendas of local and state governmental bodies as efforts were made to require them on local construction projects. The city of Cambridge has enacted a local ordinance that has put in place many of the same requirements that are found in PLAs, for all public projects. The Massachusetts legislature inserted PLA requirements in authorizing construction projects in the cities of Taunton and Brockton, and passed them over a veto by the Governor. The legislature also attempted to require PLAs on a bond authorization for the rebuilding and repair of courthouses throughout the state. Under intense negotiation between the legislature and the Governor’s office, a bill was produced that mandated PLAs for funds allocated to courthouse construction projects in Boston, Worcester, and Fall River only. The legislation created a commission to recommend establishing circumstances in which PLAs should be used. The legislation instructed the commission to consider the “appropriateness and function and the size, complexity and duration of the public construction projects” when deciding whether or not to use PLAs.5

The litigation came to a head in a 1993 Supreme Court case involving the cleanup of the Boston Harbor. In 1988, a federal court directed the Massachusetts Water Resources Authority to clean up the pollution in Boston Harbor. The Authority’s project management firm, IFC Kaiser, negotiated a PLA with the local construction unions for the project. The precedent-setting aspect of this PLA was that its use was mandated in the project’s bid specifications.6 A non-union trade group filed a lawsuit contending that requiring the PLA as a part of the bid specification violated the National Labor Relations Act. The case was appealed to the United States Supreme Court, which, in 1993, upheld the use of the PLA for the project. The Supreme Court ruling opened the door for the use of PLAs in other public Massachusetts projects, including local school construction.

School Construction Financing in Massachusetts

The School Building Assistance Program in Massachusetts has aided public school construction for more than half a century. The program began in 1948 as a three-year effort to provide resources to local communities for the building of schools for the “Baby Boom” generation, with a 25% percent reimbursement rate for the local school districts.7

The program has since grown to represent a hefty burden on state finances. After several extensions, today “the school building assistance program is the largest capital grant program operated by the Commonwealth … and the costs of the school building assistance program are increasing at an unsustainable rate.”8 In 1999, the program offered, on average, a 69% reimbursement rate for the construction and financing costs of school projects. Over the period 1991-1999 the Commonwealth of Massachusetts made total contributions to the program of more than $1.7 billion.9

The financial commitment for the state rose consistently over the 1990s. In fiscal year 1999, the annual payment for school construction projects was $201 million, a 58% increase from the $127 million appropriated in 1991.10By FY2003 school construction appropriations had jumped to $362 million, a remarkable 80% increase over the FY1999 level.11 According to the School Building Assistance Program website, for fiscal year 2003, there were 283 construction projects on the Priority List, with 19 new projects receiving authorization. The rapid growth of the program has prompted increased attention to the issue. A report entitled Reconstructing the School Building Assistance Program Policy Report, published in 2000, predicted that by FY2002 “this program will achieve ‘budget buster’ status.”12 It is within this fiscal environment that school construction costs have become an important concern in the building of public schools in Massachusetts.

III. The Evidence on PLAs

Although there is substantial anecdotal evidence that PLAs raise construction costs, no studies have provided formal statistical evidence of such an effect. To compare PLA with non-PLA costs it would be necessary to compare construction projects of a similar nature – for instance road repairs – where some projects are done with a PLA in place, and others are not. Situations such as this are rare, and even when they occur, the relevant information is difficult to obtain.

We have, however, found one suitable “natural experiment” that allows us formally to compare the costs of PLA and non-PLA projects. Driven by an increase in the student population, and encouraged by financial support from the state, many of the roughly one hundred towns and cities in the greater Boston area have financed school construction over the past several years. Some towns had PLAs in effect during the construction bidding process while others did not. Using data on construction bid costs, adjusted for inflation with an appropriate construction cost index, we were able to measure the difference in cost per square foot of construction between schools where a PLA was in effect and schools where there was no such agreement. Before reporting the results, we first need to discuss the sources of the data that we used, and explain how we adjusted for construction costs over time.

Data Sources

Our primary source of data was F.W. Dodge, McGraw-Hill Construction Information Group, a division of the McGraw-Hill Companies, in Lexington, Massachusetts. Dodge provided us with information on a variety of construction projects in the greater Boston area for the period 1995 through 2001, including the price of the winning bid, and the size of the project (in square feet). We supplemented these data with information on more recent projects by contacting school officials and construction companies, and by visiting official school web sites containing current construction-project information. Surprisingly, the Commonwealth of Massachusetts does not keep adequate or detailed information on the schools that are built largely at its expense.

We then excluded all projects with a valuation below $1 million, on the grounds that these are typically too small to be of interest to union contractors. We further focused our study on school construction projects between 20,000 and 250,000 square feet in size. Information on whether PLAs were in effect or not was obtained by contacting town and city officials (or in some cases, architects and contractors) in each of the towns for which we had cost information. Our sample comprises the 54 projects for which we had data, 31% of which involved PLAs, the remainder of which did not.13

Adjusting for Inflation

Our sample of schools covers the period 1995 to the present. In order to compare the construction costs of PLA with non-PLA schools, it was first necessary to correct for the fact that construction costs rose during this period, so that all costs could be expressed in 2001 prices. Specifically, we constructed a cost index that included both the trend in construction wages and the trend in materials costs between 1995 and 2001. Using 2001 as the base year, we first constructed a wage index, which was based on total wages and salaries for construction workers in Massachusetts, divided by the total number of workers in that sector.14

In order to account for the changes in materials costs, we constructed a price index based on the producer price index for intermediate materials, supplies, and components, as reported in The Economic Report of the President, February 2002.15 To construct the final cost index used in our analysis, we weighted the wage index and the adjusted producer price index equally to reflect the relative importance of wages and materials costs in a typical construction project. We applied the average annual increase in this index for 1995 through 2001 to estimate the expected price indices for 2002 and 2003.

Comparing PLA to Non-PLA Projects

A comparison of the key characteristics of the school construction projects in towns with a PLA (“PLA projects”) with those where there was no such agreement (“non-PLA projects”) is shown in Table 1. The most interesting finding is that PLA projects, on average, cost $35.24 more per square foot (in 2001 prices) than non-PLA projects. 

The immediate question that comes to mind is whether the cost of PLA projects is statistically significantly higher than the cost of non-PLA projects. Could it be that the difference is due to chance, so that if one were to observe more examples of projects, the difference might disappear?

One way to address the issue is with a formal regression analysis. The dependent variable is the cost per square foot of construction (in 2001 prices). The independent variable of most interest to us is a dummy variable that is set equal to 1 for PLA projects and to 0 otherwise. We control for whether the project involves new construction or a renovation by including a dummy variable set equal to 1 for new projects and to 0 otherwise. We also control for the impact of a project’s scope on the cost per square foot by controlling explicitly for square footage, and for square footage squared. This is desirable because there may be economies of scale (within reason) in school construction, so that larger schools may have lower costs per square foot. The ordinary least squares regression results are presented in Table 2.


Our results show that PLA projects add $37.88 per square foot to project costs, controlling for whether or not the project involves new construction, and controlling for the project’s square footage. A formal (one-tailed) test of the statistical significance of this coefficient gives a p-value of 0.00, which means that there is less than a 1% chance that we have accidentally found that PLA projects are more expensive than non-PLA projects. Put another way, there is at least a 99% probability that PLA projects really are more expensive than non-PLA projects. With an R 2 =0.43, the equation “explains” a respectable 43% of the variation in construction costs across towns. The equation also shows that:



  • projects involving new construction rather than renovations experience significantly lower costs per square foot; and
  • the size of a project (in square feet) has some influence on the cost per square foot. For instance, a typical 110,000 square foot project would cost $4.00 less per square foot than a typical 100,000 square foot project. This effect – lower unit costs for larger projects – applies only to projects up to 147,000 square feet in size.

IV. The Million-Dollar Question

Towns in Massachusetts are increasingly being asked to answer the question: Should Project Labor Agreements govern school construction projects? In the cases that follow, we provide some additional context for the use of PLAs in Malden, Wilmington and Milton. We then use our statistical results to estimate the savings that these towns would have obtained if they had not used PLAs. The savings are potentially large; in many cases it really is a million-dollar question.

Malden, Massachusetts

In 1996, the city of Malden began a five-year $100-million series of projects to replace its schools serving kindergarten through eighth-grade, and to remodel Malden High School. The projects were to be accomplished by closing nine existing schools, replacing five schools, and demolishing three.

On the recommendation of its construction project management firm, O’Brien-Kreitzberg, Inc., the city of Malden negotiated a PLA with the Building and Construction Trades Council of the Metropolitan District, AFL-CIO and the New Council of Carpenters, AFL-CIO. The agreement included many of the PLA provisions discussed in Section II, including: the recognition of unions as the sole and exclusive bargaining representatives of all project employees; hiring through the union referral process; the requirement of contractors to contribute to union employee benefit plans; uniform work rules and dispute resolution; and prohibiting strikes, picketing, work stoppages, slowdowns, and lockouts.16 The PLA was approved by a vote of the City of Malden municipal building committee in May of 1997; union approval followed.

In the initial phase of the project, the city bid the construction of the Beebe and Roosevelt schools as one project, with the stipulation that the project was subject to the PLA requirement. When the bids were reviewed by the city, the lowest exceeded the project budget and all bids were subsequently rejected. The project was modified and the city offered each school for bid separately.

On November 7, 1997, seven open-shop (non-union) contractors with public sector building experience filed for a motion of preliminary injunction against the use of a PLA in the bidding process. The plaintiffs argued that the PLA violated the state’s competitive bidding laws, and that they would have bid for both projects if the PLA were not included. The court denied the request for a preliminary injunction, and when the plaintiffs filed an appeal, the Massachusetts Supreme Judicial Court chose to hear the case.

The Supreme Judicial Court reaffirmed the lower court’s denial of the preliminary injunction. The court majority argued that the objectives of the state’s competitive bidding laws were to “obtain the lowest price for its work that the competition among responsible contractors [could] secure” and to create an “honest and open procedure for competition for public contracts.”17 The Court accepted the plaintiffs’ assertion that “they were inhibited from bidding, and that this inhibition could have anti-competitive effects.”18 However, the Court concluded, “that PLAs on public projects are not absolutely prohibited.”19 In echoing the decision of the Lynn case, and that of a New York case involving the restoration of the Tappan Zee Bridge, the Court stated that “the project is of such size, duration, timing, and complexity that the goals of the competitive bidding statute can not otherwise be achieved and the record demonstrates that the awarding authority undertook a careful, reasoned process to conclude that the adoption of a PLA furthered the statutory goals.”20

Interestingly, the Court did state “it may be that in certain cases, sheer size of a project warrants the adoption of a PLA. In most circumstances, the building of a school will not, in and of itself, justify the use of a PLA.” This first phase of the construction project came in on budget and on time, with no labor interruptions, according to city officials.21

According to our analysis, the new schools built in Malden could have been built more cheaply had the project not used a PLA. For example, the town would have saved $5.3 million, measured in 2001 dollars, had a PLA not been used in building the Beebe School.

Wilmington, Massachusetts

Wilmington is a suburban industrial town that sits on the watershed of the Ipswich River some eighteen miles north-northwest of Boston. Faced with an increasing birth rate between 1986 and 1995, and a subsequent school enrollment surge in the 1990s, the Wilmington School District began to suffer from overcrowding. School enrollment increased 32% (900 students) in the decade from 1989, and another 9% (300 students) from 1998 to 2000.22 Under the pressure of this increase, and with an eye towards future growth, in the spring of 1997 the town voters approved a Proposition 2 1/2 override to fund a new $23 million middle school. Proposition 2 1/2 was a Massachusetts ballot initiative in 1980 that capped the rate of taxation on real and personal property and also limited the annual increase in the property tax levy. Wilmington’s residents decided to accept higher taxes in return for less-crowded schools.

The town selected Architectural Resources of Cambridge to design the new middle school. The town bid the project with a PLA requirement for the construction project. Construction began in December of 1998 and the project was completed as scheduled in time for the start of the 2000 school year. The citizens of Wilmington also spent more than necessary for their new Middle School. Our statistical analysis indicates that, in the absence of a PLA, Wilmington would have saved $5.7 million, measured in 2001 dollars, on constructing their new Middle School.

Milton, Massachusetts

Milton has recently embarked on an ambitious $100 million program of school construction. The project includes creating a new high school ($50.3 million) out of the existing middle school, a new middle school ($28.6 million), the Glover school ($12.5 million) and the Collicot and Cunningham school ($27.3 million). The town has signed a Project Labor Agreement, so contractors must follow PLA rules. The bids for the new High School and the Glover School portions of the project came in substantial higher than the budget estimates of the town school committee. The lowest bid for the High School was $4 million over the school committee estimate, while the Glover School bids came in $800,000 over the estimates.23 Our calculations show that Milton would save $15.3 million on the construction of these four schools if it did not have a PLA; of this, $6.6 million would be saved on the construction of the high school alone.

V. Conclusion

It is widely believed that Project Labor Agreements add to the cost of construction projects. However, there has until now been no statistically sound study of whether PLAs add to construction costs in practice, and if they do, how large the effect is. By carefully constructing a database with information on the costs of school construction projects undertaken in the greater Boston area since 1995, and comparing the costs in towns with, and without, PLAs, we found that:

  1. PLA projects have higher construction costs. We are more than 99% confident of this assertion, based on the available data.
  2. PLA projects add an estimated $37.88 extra per square foot of construction (in 2001 prices), representing a 22.1% increase in costs for the average PLA project.

Applying our results to the experiences of three Massachusetts communities, we find that each of these towns would have lowered its construction costs by over $5 million in the absence of a PLA. Our study has identified seventeen schools that were put out to bid under PLA agreements in the greater Boston area since 1997; this is as complete a list as we were able to compile. Our estimates show that the cost of building these schools was, in total, $85.5 million (in 2001 prices) higher than it would have been if PLA agreements had not been put in place. About 70% of this additional cost was borne by the Commonwealth, with the remainder paid for by the towns themselves. When Massachusetts communities are asked whether to approve Project Labor Agreements in the future, they will now know that they are facing a multi-million-dollar question. The issue is likely to arise with increasing frequency in the near future, as Boston’s “Big Dig” winds down and the released workers seek other work more aggressively.

* Jonathan Haughton, Ph.D., is an Associate Professor in the Economics Department at Suffolk University and a Senior Economist at the Beacon Hill Institute (BHI) for Public Policy Research at Suffolk University. He holds a Doctorate from Havard University. Darlene C. Chisholm is an Associate Professor in Economics and a Senior Economist at BHI. She holds a Doctorate from the University of Washington. Paul Bachman, MSIE, is a Research Assistant at BHI. He holds a Master of Science in International Economics from Suffolk University. The authors would like to thank Dali Jing, Corina Murg and Hatesh Radia for their contributions to this report.

The Beacon Hill Institute at Suffolk University in Boston focuses on federal, state and local economic policies as they affect citizens and businesses. The institute conducts research and educational programs to provide timely, concise and readable analyses that help voters, policymakers and opinion leaders understand today’s leading public policy issues. This BHI report, originally published in March 2003, is republished herein by permission. 


1 "United States General Accounting Office, “Project Labor Agreements: The Extent of Their Use and Related Information,” (May 1998). 

2 Ibid.

3 Ibid. 

4 Worcester Municipal Research Bureau, “Project Labor Agreements on Public Construction Projects: The Case For and Against,” Report No. 01-4, May 21, p. 7, (2001).

5 Herbert R. Northrup and Linda E. Alario, “Government-Mandated Project Labor Agreements in Construction, The Institutional Facts and Issues and Key Litigation: Moving Toward Union Monopoly on Federal and State Financed Projects.” Government Union Review, vol. 19 number 3 p. 91 (2000). 

6 Ibid., pp. 12-13.

7 Massachusetts Executive Office of Administration and Finance, Reconstructing the School Building Assistance Program, Policy Report Series No. 3, January, 2000. 

8 General Laws Of Massachusetts, Chapter 70: “School Funds and State Aid for Public Schools.” 

9 Massachusetts Executive Office of Administration and Finance, Reconstructing the School Building Assistance Program, p. 1.

10 Fiscal year 1999 refers to the period July 1, 1998 through June 30, 1999. 


12 Massachusetts Executive Office of Administration and Finance, Reconstructing the School Building Assistance Program, p. 1. 

13 PLA contracts were in effect in the following towns: Boston, Lawrence, Lynn, Malden, Melrose, Milton, Waltham and Wilmington. 

14 The source of wage and salary data is the Bureau of Economic Analysis; the source of the total number of workers is the Bureau of Labor Statistics. 

15 The source of the producer price index is Table B-66, “Producer Price Indexes by Stage of Processing, Special Groups, 1974-2001,” The Economic Report of the President, February 2002. 

16 John T. Callahan &Sons, Inc. &others vs. City of Malden &another. SJC-07959, Lexis-Nexis 493, July 22, p. 2 (1999). 

17 Ibid., p. 2.

18 Ibid., p. 2.

19 Ibid., p. 2.

20 Ibid., p. 6. 

21 Worcester Municipal Research Bureau, p. 9. 

22 Town of Wilmington Master Plan, 2001, pp. 181-183. 

23 Kimberly Atkins, “Milton May Reject School Bids," Boston Globe, May 2, 2002, Globe South,p 2.



at $130 per hour is now cheaper than the $140 one plumber-three helper combination. This same reversal results even if the plumber wage is raised to $55: the two plumber-one helper combination now costs $140 per hour, and the one plumber-three helper combination costs $145. In the last example, the relatively greater wage increase for helpers reduced their employment as compared with that of plumbers. Under either minimum wage scenario, the employment of plumbers has increased at the expense of the employment of helpers.

Because minorities are disproportionately represented among the relatively unskilled laborers, this shift to higher skilled labor accompanying increased wage rates should adversely affect the employment of minority construction workers. In 1990, 32.0 percent of the 1.26 million construction laborers (including helpers and apprentices), but only 22.2 percent of all 5.71 million nonsupervisory construction workers, were minorities. Blacks constituted 13.0 percent of laborers and 8.5 percent of all construction workers; Hispanics, 16.7 percent of laborers and 11.7 percent of all construction workers.

Minority employment may also be affected by unionization, apart from union-based higher wages. Minorities were traditionally viewed as less likely than whites to be union members because of: (1) less access to information about vacancies in unions in part because of fewer friends, relatives, and neighbors with union experience; (2) poorer training as youngsters and therefore lower qualifications to enter union apprenticeship programs; and (3) active discrimination on the part of union officials. Recent affirmative action efforts on the part of unions have countered these tendencies.

Indeed, aggregated data for the years 1986-1994 reveal that 23.8 percent of all, 24.0 percent of white, and 21.8 percent of minority nonsupervisory construction workers in the construction industry were union members. Part of this mild racial disparity results from the lower unionization rate for laborers, who, as noted above, have a relatively high representation of minorities. The unionization rate was 20.3 percent for laborers (19.9 percent white; 22.1 percent minority), but 25.0 percent for the remainder of nonsupervisory construction workers (25.3 percent, white; 21.7 percent, minority). In contrast, Ashenfelter (1972, p. 451) notes that, in 1967, more than half of white and about a quarter of black skilled construction workers, plus 28 percent of white and 35 percent of black laborers, were unionized. These data illustrate that the pronounced narrowing of the racial gap in construction unionization rates in the 1970s and 1980s was produced mainly by a sharp reduction in the union membership rate of skilled whites.

If individuals interested in the relatively well-paying construction jobs but not employed in construction—those employed in lower-paying jobs, the unemployed, and nonparticipants in the labor force—are disproportionately minority, then these measured unionization rates understate the racial unionization gaps based on this expanded supply of potential construction labor. Also, given the many allegations of hiring hall discrimination in referred hours of work, construction union members not employed in the building trades may be disproportionately minority (unions typically allow their nonworking members to take jobs in other industries). In any event, because only one of four construction workers is a union member, much of the potential employment effects of higher wage rates are likely to occur in the nonunion sector.

The scale and substitution effects discussed above also apply to the skill mix of construction workers. Thus, higher wages should be negatively correlated with the relative employment of laborers as opposed to skilled workers. Furthermore, increased wage rates likely reduce total employment in the construction industry. Unionization and construction employment are negatively correlated across metropolitan areas, in part because of the very strong positive correlation between unionization and wage rates. Non-wage effects of unions on construction employment may be either positive (if, for example, unions increase construction activity by lobbying successfully to replace publicly-owned structures that could have lasted years more) or negative (if the presence or growth of unions discourages construction activity).

Empirical Results

My primary regression model implies that a $1 increase in construction hourly wage rates would induce a 146,000 nationwide job loss in construction occupations—39,000 laborers and 107,000 skilled workers. This reduction in jobs represents 110,000 lost jobs for minorities (28,500 laborers; 81,500 skilled) and 36,000 lost jobs for whites (10,500 laborers; 25,500 skilled).

These empirical results, in conjunction with a conservative estimate that the impact of Davis-Bacon is equivalent to a 25 cent increase in the overall construction mean hourly wage, imply that elimination of the Davis-Bacon Act would increase the total number of construction jobs nationally by more than 36,000. This increase would reflect a disproportionate gain of 27,000 jobs for minorities (7,000 laborers; 20,000 skilled), plus a gain of 9,000 jobs for whites (3,000 laborers; 6,000 skilled). Economy-wide employment increases are lower than those for the construction trades because some additional construction jobs would represent shifts from other occupations.

Because skilled workers outnumber laborers by more than 3.5:1, laborers would benefit disproportionately from repeal. Thus, in addition to removing the inefficiencies induced by wage floors, repealing Davis-Bacon would increase the relative construction employment of minorities and laborers. The disproportionate gains for minorities and laborers reflect the considerable positive equity impact of Davis-Bacon repeal. A complete equity analysis must also take into account the wage increases enjoyed by minorities and laborers employed on Davis-Bacon projects. Repeal of state prevailing wage laws would be expected to produce similar effects.

These employment estimates somewhat understate the expected increase in construction jobs resulting from Davis-Bacon repeal. First, I have used a conservative estimate of the wage effect. Second, the Davis-Bacon Act affects employment in Census occupational classifications other than construction. Third, repeal of Davis-Bacon could considerably weaken unions and induce a further reduction of construction wage rates.

IV. Conclusions

Higher wage markets are associated with lower construction employment and, most striking, substantial reallocations of employment across race and skill groups, to the detriment of minorities and laborers. These wage effects imply that repealing Davis-Bacon and state prevailing wage laws would have particularly salutary effects on construction employment opportunities for minorities and laborers.

Research on this issue can be continued along at least two avenues: study of the employment effects of state prevailing wage laws, and replication of my study with the soon-to-be-forthcoming Census 2000 data. Differential effects across time and space might be correlated with the extent of discrimination, suggesting a decline in racial redistributive effects as discrimination diminishes over time, and as opportunities increase for both full-time construction employment and secondary occupations for construction workers.




Allen, Steven G. “Much Ado About Davis-Bacon: A Critical Review.” Journal of Law and Economics, October 1983, 26(3), pp. 707-36.

________. “Can Union Labor Ever Cost Less?” Quarterly Journal of Economics, May 1987, 102(2), pp. 347-73.

Ashenfelter, Orley. “Racial Discrimination and Trade Unions.” Journal of Political Economy, May-June 1972, 80(3), pp. 435-64.

Bloch, Farrell. “Minority Employment in the Construction Trades.” Journal of Labor Research, Spring 2003, 24(3), pp. 271-91.

Goldfarb, Robert S. and Morrall, John F. III. “The Davis-Bacon Act: An Appraisal of Recent Studies.” Industrial and Labor Relations Review, January 1981, 34(2), pp. 191-206.

Marshall, F. Ray, Cartter, Allan M., and King, Allan G. Labor Economics: Wages, Employment, and Trade Unionism (3rd ed.). Homewood, Ill: Richard D. Irwin, 1976.

Metzger, Michael R. and Goldfarb, Robert S. “Do Davis-Bacon Minimum Wages Raise Product Quality?” Journal of Labor Research, Summer 1983, 4(3), pp. 265-72.

Thieblot, Armand J., Jr. The Davis-Bacon Act. Philadelphia: University of Pennsylvania Press, 1975.

U.S. General Accounting Office. The Davis-Bacon Act Should Be Repealed. Washington: U.S. General Accounting Office, April 27, 1979.

PREVAILING WAGES: COSTLY TO STATE AND LOCAL TAXPAYERS by Frank Gamrat, Ph.D. Allegheny Institute for Public Policy*

Key Findings

Prevailing wage laws exist at both the federal and state levels and are responsible for increasing the cost of government. The general purpose of a prevailing wage law is to artificially raise the wages of workers participating in government construction projects. Michigan, Florida and Ohio have experienced savings of 10 percent and higher on construction in the absence of a prevailing wage law. In Pennsylvania, prevailing wage laws apply to construction at all levels of government. This report shows that Pennsylvania construction savings would be at least 10 percent based on the difference between the prevailing wages set by the state and true market wages. Based on that finding, a number of observations can be made.

  • For the fiscal year 1999, the Commonwealth had $2.12 billion worth of construction expenditures. Absent the prevailing wage law, the state government could have saved $212 million on construction in FY99. That amounts to 2.4% of the individual income tax collected that year.
  • Local governments within Pennsylvania spent an additional $3.48 billion on construction in the same year. Again without the prevailing wage law, $348 million could have been saved. Money which could have been used to lessen the burden of property taxes by 3.6% ($9.66 billion collected in 1999).
  • School districts and the Commonwealth commit more than a half a billion dollars per year to school construction. In 1998 there were 66 school building projects, eligible for state reimbursement with a total value of over $730.6 million. With no prevailing wage requirement, school construction alone would have saved more than $73 million.

Beyond the monetary costs imposed on taxpayers, prevailing wage laws impose strict craft-based classifications and restrictive apprenticeships regulations. This limits employment opportunities in construction for entry level, or low skilled workers.

  • During the period its law was set aside by a federal judge, Michigan experienced an annual construction job growth rate that was almost 4.5 times the rate of growth as when the law was in effect (4,000 vs. 17,600). Even when controlled for external factors such as weather, the base result is still the same: More construction jobs were created during the period of no prevailing wage requirement, than when the law was in force.
  • During the 30 months without Michigan's prevailing wage law, there were 116.3 construction jobs created per 1,000 overall jobs, compared to the 78.6 per 1,000 jobs in the 30 month period before the repeal.

Introduction: The Cost of Prevailing Wage Laws

In 1931 the federal government passed the Davis-Bacon Act which requires that all contractors working on federal projects pay their employees no less than prevailing wages. According to the Davis-Bacon Act, prevailing wages are to be paid to any employee working on a federal government project with a value of $2,000 or more.

Prevailing wages were originally intended to increase the wages of local laborers and protect them from lower-wage migrants. At the time supporters of this legislation claimed that workers who were paid more would help spend the country out of the Great Depression. This argument gained momentum and eventually filtered its way down to the state level as over 30 states followed suit and passed "little Davis-Bacon" laws. The unintended consequence of the law was to force non-union contractors to pay their workers union-scale wages. The end result of the federal and state level Davis-Bacon laws is to raise the cost of labor and subsequently the cost of government projects which are ultimately passed on to the taxpayers.

The Federal Davis-Bacon Act

Although the U.S. Department of Labor sets the wage, it is almost invariably higher than what would prevail in the free market. According to the Department of Labor, the wage is determined by collecting wage data through the voluntary submission of wages by contractors, labor organizations, public officials, and other interested parties. According to the Congressional Budget Office (CBO), "those procedures (of the Department of Labor) as well as the classification of workers who receive prevailing wages, favor union wage rates...."1

The CBO in its Budget Options for 2001, recognizes that the federal Davis-Bacon law inflates the cost of government. They estimate that if Congress were to repeal the law, the federal government could save $9.5 billion from 2002 to 2011.2 As is stated in the report, "Repealing the Davis-Bacon Act would allow the federal government to spend less on construction...In addition, it would probably increase the opportunities for employment that federal projects would offer to less skilled workers."3 In absence of repeal, the CBO suggests raising the threshold from $2,000 to $1 million. By raising the threshold, the CBO estimates that the federal government could save $1.3 billion in federal outlays over the 2002 to 2011 period.4

Even though many groups support the repeal of the Davis-Bacon Act, there are other special interest groups that are steadfast in its support. Davis-Bacon gives union contractors an advantage over non-union contractors in the bidding for government projects. Davis-Bacon not only dictates the wages that must be paid to workers on government contracts, but it also dictates the hourly price for "fringe" benefits.5 Fringes for union workers are programs that are paid from trusts that have been built from dues payments and are not subject to payroll taxation. However, for the non-union firm the absence of such programs means that fringes must be paid directly to the employee as a supplement to the hourly wage and thus subject to payroll taxes. Therefore, not only are non-union firms required to meet the wage being paid by union firms, but must exceed them through fringe payments and then must pay more in payroll taxes than their union counterparts. As a result many non-union contractors pass on government projects, further biasing upwards the cost of construction.

The Act not only raises the cost of labor, which many non-union firms cannot match, but it also imposes rigid craft-based job classifications and restrictive apprenticeship regulations. This sharply limits an employer's ability to hire and train unskilled workers. In many cases these unskilled workers are minorities, which have historically been kept out of trade unions. It is believed that racism was a primary motivating factor in passing the Davis-Bacon Act. One purpose of the Act was to prevent contractors from using African-Americans from Southern states on projects in predominantly white Northern states. As one congressman stated in 1931, it prevents contractors from using "cheap colored labor."6

Nationally, black workers account for approximately 10.4% of the total workforce. However in the construction industry, blacks account for only 6.4% of all workers.7 Moreover, they have historically been shut out of unions. And since many black contractors are often not as well capitalized as large union shops, they are unable to pay prevailing wages and thus unable to bid on government projects. Many do not even try. Therefore the racial bias that was a part of the impetus of the Davis-Bacon Act of 1931, still echoes 70 years later.

Davis-Bacon at the State Level

As mentioned above, after the passage of the Davis-Bacon Act at the federal level, 32 states and the District of Columbia enacted what were to be known as "little Davis-Bacon Acts".8 As of 2000, 31 states and the District of Columbia have prevailing wage laws on the books. Nineteen states do not have such a law. Table 2 in the Appendix, provides details such as the year the law went into effect, the year of repeal, where applicable, and effective rate chosen.

The nation's first prevailing wage law was enacted by Kansas in 1891 as an "eight hour day law". This law focused mostly on prohibiting overtime except in an emergency."9 The last state level prevailing wage law was passed in Minnesota in 1973. Of the 19 states that do not currently have prevailing wage laws, 10 repealed prior Davis-Bacon laws (from 1979 through 1995.) Four arguments have led to the abandonment of prevailing wage laws: laws force employers to pay more for labor than the market would otherwise dictate; it allows employers to discriminate in hiring workers; it raises the cost of government; it increases administrative costs.

For example, to ease the burden on school districts, the Florida legislature dropped the prevailing wage requirement from school construction from 1974 through 1978. They found that the average yearly savings on school construction were about 15% ($37 million). After learning of the savings, the legislature repealed the entire prevailing wage law in 1979.10

Each state has its own process by which it determines prevailing wages. Some states set their wages through an elaborate process or through the local collective bargaining agreements between unions and contractors. Fifteen of the 31 states with a prevailing wage law use local collective bargaining rates. Some states simply choose to use federal rates and classifications. Below we compare four states, including Pennsylvania, as to how prevailing wage works.


Oregon has had prevailing wage law since 1959. The Oregon Prevailing Wage Law (PWL) defines the rate to be paid as the rate of hourly wage, including fringe benefits paid in the locality to the majority of workers employed on projects of similar character in the same trade or occupation. The wording of Oregon's PWL may lead one to believe that local wages were recorded and used as the standard on government projects. However, according to the Oregon Bureau of Labor and Industries, prior to 1995, no survey of wages was ever taken. Instead, Oregon either used the federal Davis-Bacon rate or the collective bargaining rates of local unions.

Oregon conducted its first survey of contractors registered to perform heavy, highway, and commercial construction in 1995. The results substantiate the notion that unions push the wage rate above what would prevail in a free market. When the Oregon Bureau of Labor and Industries compared all occupations, they found that the average hourly wage in their most unionized area (Clatsop, Columbia, and Tillamook counties) were 27.6% higher ($18.49 vs. $14.48) than its least unionized area (Crook, Deschutes, and Jefferson counties). However an even larger gap occurred in the payment of fringe benefits. The average fringe benefit rate per hour in the two areas differed by 140% ($4.73 vs. $1.97). As mentioned above union shops have the advantage in fringe benefits because they are paid from well funded trusts whereas non-union contractors usually pay them as monetary supplements making them subject to payroll taxes which the non-union firm must pay. The "total" hourly union wage is 41% higher ($23.22 vs. $16.45).

Comparing occupations by union and non-union shops yields similar results. For example, painters, paperhangers, plasterers and stucco masons had an average hourly rate in the unionized zones of $15.56, while the average rate in non-unionized zones is $10.12.11 The average fringe benefit rate per hour is $3.26 in the union areas, but only $0.24 in the non-union areas.


In 1997, Ohio passed a revision to their prevailing wage law which allows school districts the option of requiring prevailing wages to be paid on school construction projects.12 This revision is to be a five year experiment by the Ohio legislature to see if school districts are saving money, receiving quality work on school building construction, and what are the impacts on the wages of construction workers on school projects. The Ohio Legislative Budge Office (LBO) commissioned a first year study of the experiment. The results of the study show that where there are savings, the average rate is 6.1% (not counting administrative costs). There has been no appreciable decrease in the quality of school construction. Employment in the construction industry continued to grow despite the exemption of school construction from the prevailing wage law. Although causality cannot be determined, average hourly rates continued to grow despite the exemption. However, since the prevailing wage requirement was optional, some districts, in mostly heavily unionized areas, held onto the requirement.

Ohio's Legislative Budget Office surveyed the 611 school districts (396 responded--65%) after the first year of implementation of the option. Of the 396 responding districts, 14 (4%) still required their contractors pay a prevailing wage while 320 (81%) did not.13 Some of the respondents (37%) believed that by eliminating the requirement they would be able to save money on school construction. Others saw the change as an opportunity to give local contractors a chance to do some work. Some respondents expressed concern about the quality of work performed as they equated higher wages with better workmanship. However, most respondents noted that the elimination of the prevailing wage requirement did not adversely affect quality. The report finds that the skill level and quality of workmanship do not increase by simply requiring and paying prevailing wages.14

The Ohio Legislative Budge Office also surveyed contractors who bid on school projects. The LBO asked each responding contractor to submit two bids: one with a prevailing wage and one without. The difference was then used to estimate savings. 379 contractors responded to the survey. For 136 contractors (36%) there was no difference between their bids. Many of these respondents indicated that they were union shops. The author then notes that even though there were no savings in the contract bid, there should still be savings due to reduced compliance and administration costs. Two of the respondents (1.5%) noted that without the prevailing wage requirements, their bid price would be higher due to the inefficiencies in using unskilled labor. They reasoned that the inefficiencies would translate into longer hours to complete any project. However, 241 of the contractors (64%) indicated that they would bid the job lower without a prevailing wage requirement. The average savings rate between the two bids is 10.2%.

The LBO also noticed that savings depended upon the location, rural or urban, of the school district. In urban districts, where there were savings, the average rate was 9.4%. However, in rural districts, where there were savings, the average rate was 14.4%. One explanation is that since urban wages are typically higher than those found in rural communities (due to tighter labor markets), prevailing wages are often "imported" into the rural region. Removing the prevailing wage requirement allows rural contractors to do more work at a lower wage. The survey found that savings by trade are larger in rural counties than in urban counties. This again is a reflection on the tightness of the labor market lessening the difference between the union and non-union wage in an urban versus a rural setting.

The conclusion of the Ohio study notes that "where there are savings, the savings may be significant."15 However, the savings do depend upon the location of the district and the labor climate in which they operate. The Ohio LBO also cautions that extrapolating those savings across all construction projects may be premature since school construction only accounts for 5% of all construction in Ohio and some contractors may simply have passed on these projects since they are such a small percentage of their business.


In December 1994, Michigan's prevailing wage law was found to be invalid by a federal judge. The judge ruled that the state's prevailing wage law was preempted by ERISA, a federal pension law. This decision was later reversed by an appellate court in June 1997. As a result Michigan's prevailing wage law was not enforced for a period of 30 months. Richard Vedder studied how the hiatus effected construction employment and government spending.16

With regard to construction job growth, during the break, Michigan experienced an annual growth rate that was almost 4.5 times the rate of growth as when the law was in effect (4,000 vs. 17,600). Even when controlled for factors such as weather, seasonality of construction, and business cycles, the author finds the base result is still the same: More construction jobs were created during the period of no prevailing wage requirement, than when the law was in force. During the 30 months without the law, there were 116.3 construction jobs created per 1,000 overall jobs, compared to the 78.6 per 1,000 jobs in the 30 month period before the repeal.

Prevailing wages have been shown above to increase the cost of public construction. But, how much so depends upon three factors: By how much does the prevailing wage exceed the free market wage; what percentage of the total construction costs are labor; and what impact on productivity does the prevailing wage have.

In Michigan, Vedder finds that the prevailing wage was 30-40% more than the market wage. He also estimates that labor costs are, on average, about 25% of the value of a construction contract. He also finds that there is no reliable evidence that labor productivity changed in the absence of prevailing wages.17 Therefore, with prevailing wages that are as much as 40% higher than free market wages, and labor comprising 25% of overall construction costs, the cost of a contract will be raised by approximately 10%. Evidence from school construction that took place during the 30 month reprieve shows that savings were over 10% (13-16%). This could actually be due to improved managerial control over workforce and thus lower total overhead costs and compliance.

In fiscal 1995, Michigan state and local governments spent almost $2.51 billion on construction outlays. Assuming a savings rate of 10%, the elimination of the prevailing wage could have saved them $251 million. Adding to that total, non-construction outlays that are subject to prevailing wages, the overall savings to Michigan was about $275 million in 1995.18 As noted by Vedder, "It is the equivalent of slightly over five percent of the revenue raised by the Michigan individual income tax in fiscal 1995 ($5.473 billion). "(R)epealing Michigan' prevailing wage law would have an impact the equivalent of giving every taxpayer a rebate equal to five percent of his state income tax payments."19


Pennsylvania instituted its prevailing wage law with the Pennsylvania Prevailing Wage Act of 1961.20This Act applies prevailing wages to any publicly funded construction project with a value of $25,000 or more. (In 2001, a bill was introduced in the State Senate (SB 821) which would have raised the threshold to $500,000. This bill, like previous attempts to amend or repeal the prevailing wage act, did not succeed.) Pennsylvania's prevailing wage act also applies to established trades outside of the construction industry, such as printing contracts at the state level (local leaders have the option on printing contracts) and highway construction. Maintenance work is not covered, however reconstruction work is, causing many court battles seeking to define when a project is "maintenance" or "reconstruction".

When determining the rates to be used, the Act gives the Secretary of Labor the power to use the local collective bargaining rate as the prevailing wage. However, the Ridge Administration commissioned a Prevailing Wage Advisory Board to construct county-by-county surveys to set and update new rates. In 1997, they claimed that by establishing new rates, taxpayers should save $100 million in public construction costs. They claim that "the Prevailing Wage Law will now be able to function as intended because the Ridge Administration has taken the initiative to determine rates which truly reflect local wages on construction projects in each of our counties."21

Government Construction Expenditures in Pennsylvania

For the fiscal year of 1999, the Commonwealth of Pennsylvania had $2.12 billion worth of expenditures on construction.22 It will be assumed that all of the projects (given the low threshold of $25,000) were subject to the state's prevailing wage law.23 Using conservative estimates from the Michigan experience, it can be assumed that the state's taxpayers could have saved 10% of these costs in the absence of prevailing wages without any noticeable differences in workmanship. This amounts to $212 million for FY99, which represents 2.4% of the individual income tax collected in 1999 ($8.85 billion). Local governments within Pennsylvania spent an additional $3.48 billion on construction in the same year. Again, assuming that 10% could have been saved without prevailing wages, this amounts to $348 million which could have been used to lessen the burden of property taxes by 3.6% ($9.66 billion collected in 1999).

In the Pittsburgh region, the differences in the average free market wage and the prevailing wage is shown in Table 1. Table 1 shows a sampling of occupations that are covered by the Prevailing Wage Act of 1961. The first column gives the average free market wage in the Pittsburgh MSA while the second column gives the corresponding prevailing wage for the same occupation. The third column shows the difference between the two. The largest difference occurs with Electric Linemen ($10.03), while the smallest occurs with Cement Masons/Finishers ($2.85). The average difference is just over $6.00 per hour.

The real discrepancy happens when fringe benefits are added to the mix. Column 5 lists the hourly fringe benefits as required by the prevailing wage. The average hourly fringe benefit is $8.57, with a range of $6.12 (Landscape Laborers) to $11.65 (Sheet Metal Workers). As mentioned above, union contractors pay fringe benefits through programs that are run from trusts and are not subject to payroll taxation. However, for the non-union firm the absence of such programs means that fringes must be paid directly to the employee as a supplement to the hourly wage and thus subject to payroll taxes. When adding the cost of the fringe benefits to the hourly wage, the average payment that is to be made to a worker under the prevailing wage law is $31.20. In the free market, hourly fringe benefits can be conservatively estimated at 30% of the hourly wage. Thus the free market fringe is approximately $5.00 per hour, bringing the total wage plus benefits to $21.62.

According to the Pennsylvania Department of Education, school districts and the Commonwealth commit more than a half a billion dollars per year to school construction. In 1998 there were 66 school building projects, eligible for state reimbursement with a total value of over $730.6 million.24 Assuming that Pennsylvania experiences the same savings rate that was found in both Ohio and Michigan (10%), savings would exceed $73 million. This represents only a fraction of all school projects that are undertaken that would fall under the guise of the prevailing wage act. In the first two months of 2001, there were 169 prevailing wage projects in Allegheny County that were submitted to the state's Bureau of Labor Law Compliance.25 Of these, 76 or 45% of the projects were school related. If Pennsylvania were to follow the lead of neighboring Ohio, and allow school districts to be exempted from the prevailing wage law, savings to individual school districts could be substantial.


In 1931, the Davis-Bacon Act was passed with the promotion of providing a "level playing field" for local contractors bidding on lucrative government contracts. The idea of paying workers more, or giving them "fair" wages, quickly spread throughout state level governments as over 30 states passed similar laws. However its roots were firmly planted in racism in an attempt to prevent contractors from using "cheap colored labor." The intended consequence of forcing non-union contractors into paying their employees union scales wages has cost taxpayers billions of dollars nationwide each year.

The Congressional Budget Office recommends either eliminating or modifying the Davis-Bacon Act. By repealing Davis-Bacon, the CBO estimates that the federal government can save $9.5 billion over the next ten years. In the absence of repeal, the CBO advocates raising the threshold from its current level of $2,000 to $1million. This would save the federal government approximately $1.3 billion over the next ten years.

The discrepancies caused by prevailing wages at the state level are well documented. Empirical evidence from Oregon, Michigan and Pennsylvania show that prevailing wages are on average 25-40% higher than free market wages. The real difference occurs with fringe benefits, which in Oregon were shown to be nearly twice what was being offered in the private sector. Fringes compound the problem by adding additional tax burdens onto non-union contractors. With prevailing wages higher than free market wages, many non-union contractors simply pass on government projects. This leads to less competition and higher costs for government construction which are ultimately borne by the taxpayer.

Making the prevailing wage law an option at the school district level, has resulted in substantial savings for Florida (which ultimately led to the statewide repeal), Ohio and Michigan. Average savings were about 10%. If Pennsylvania were to follow in the footsteps of these states and make the prevailing wage an option at the school district level, the Commonwealth could see annual savings of $73 million.

Prevailing wage laws not only cost taxpayers more in construction costs, it also limits job creation. During the 30 month break in Michigan's prevailing wage law, job creation in the construction industry per 1,000 jobs increased from 79 to 116 (47%). In addition, it sets strict guidelines on job classifications as well as the number of apprentices and trainees that may be used on a job. In many cases these unskilled workers are minorities, which have historically been kept out of trade unions.

The Davis-Bacon Act of 1931 and the subsequent "Little Davis-Bacon Laws" that have sprung up at the state level, are firmly based in racism and are very costly to taxpayers. The repeal of these laws would further increase the opportunities for unskilled workers to enter the construction industry and earn the decent wage originally promised by the legislation. Its removal would also put more money into the pockets of taxpayers while decreasing the cost of government.

This report, Allegheny Institute Report #02-02, was first published by the Allegheny Institute in February 2002. © by Allegheny Institute for Public Policy. All rights reserved. Note: Nothing written here is to be construed as an attempt to aid or to hinder the passage of any bill before the Pennsylvania General Assembly.

Frank Gamrat, Ph.D., is a Senior Research Associate at the Allegheny Institute for Public Policy. Address: 305 Mt. Lebanon Blvd., Suite 305, Pittsburgh, PA 15234. Phone: 412-440-0079. Fax: 412-440-0085. Web site:

1 Congressional Budget Office. Budget Options for 2001. February 2001. Section 920-05-A. Page 7.

2 Ibid.

3 Ibid.

4 Congressional Budget Office. Budget Options for 2001. February 2001. Section 920-05-B. Page 8.

5 Fringe benefits were added to the Act in 1964 and include medical care, workman's compensation, pensions, vacation pay, etc.

6 Vedder, Richard. "Michigan's Prevailing Wage Law and Its Effects on Government Spending and Construction Employment." Mackinac Center for Public Policy. 1999. Page 4.

7 Wilson, Mark and Rebecca Lukens. "Four Reasons Why Congress Should Repeal Davis Bacon". The Heritage Foundation Backgrounder. June 7, 1995. No. 252.

8 9 states already had such laws on the books and 9 never had such a law.

1 Thieblot, Armand J. Jr., "Prevailing Wage Legislation: The Davis-Bacon Act, State "Little Davis Bacon" Acts, the Walsh-Healey Act, and the Service Contract Act." University of Pennsylvania. The Wharton School Industrial Research Unit. 1986.

10 Broward County Florida installed their own prevailing wage law in 1981.

11 An area or zone is classified as unionized if 40% or more of the workers belong to a union.

12 Ohio Senate Bill 102.

13 The remaining 72 had not adopted a formal policy.

14 Lundell, Allan. "A Study of the Effects of the Exemption of School Construction and Renovation Projects from Ohio's Prevailing Wage Law." Ohio Legislative Budget Office. September 1998. Page 7.

15 Ibid. Page 13.

16 Vedder, Richard. "Michigan's Prevailing Wage Law and Its Effects on Government Spending and Construction Employment." Mackinac Center for Public Policy. 1999.

17 Ibid. Page 14.

18 Ibid. Page 14.

19 Ibid. Page 15.

20 Pennsylvania Prevailing Wage Act of 1961, P.L. 987, No.442.

21 News Brief from the Office of Governor Tom Ridge. "Taxpayers to Save Millions Through New Prevailing Wage Rates." Vol. 2, Issue 6. February 28, 1997.

22 U.S. Census Bureau. Pennsylvania State and Local Government Finances by Level of Government: 1998-99.

23 Does not include expenditures on highway construction or printing contracts which are both covered by prevailing wages.

24 Dept. of Education must approve plans and specifications for all public school construction or reconstruction, and for ordinary repairs or maintenance work for any second, third, or fourth class district.

25 Determination dates were for January and February of 2001 with award dates through June 1st.



In 1931, in the midst of the Great Depression, the United States Congress passed the Davis-Bacon Act, a law that requires contractors to pay "prevailing wages" on construction projects undertaken for the federal government. This legislation led to the passage of "little Davis-Bacon Acts," or "prevailing wage" laws, in over 40 states, including Kentucky.

What are prevailing wages? The short answer is that in many jurisdictions, including the federal government, prevailing wages are typically wages set at or near the union-scale level. Prevailing wage laws, then, force contractors on government construction or other projects to pay their employees at the same rate as unionized members of the relevant occupation--whether it be bricklayers, carpenters, electricians, or other categories of workers--even if non-union contractors could perform the same work less expensively by paying their workers lower but mutually agreed-upon wages.

Ever since the enactment of the Davis-Bacon legislation, the question of the impact of prevailing wages on the market for construction worker labor has been an issue among economists. The major question raised is simple enough: Does the enforcement of the prevailing wage statutes distort labor market outcomes associated with public construction projects? The report under review here, prepared by the Kentucky Legislative Research Commission, deals with this issue in the particular case of the Commonwealth of Kentucky, which has had its own prevailing wage law since 1940. Currently, the Kentucky law is applicable for construction projects with a total cost of $250,000 or more. In its letter of transmittal (memorandum) to the Governor of Kentucky, the Commission describes the matters dealt with as consisting of "whether or not Kentucky's prevailing wages accurately represented local wages and if the use of prevailing wages increased the costs of public construction" (p. iii).

By way of background, two different procedures are employed in determining prevailing wages in Kentucky. For 81 of the State's 120 counties, the prevailing wages are set by an entity known as the Kentucky Labor Cabinet. These rates are based on wage data collected from contractors and unions at public hearings held in the various localities. The rule for determining the prevailing wage is straightforward. If the submitted data show that a majority (51%) of workers in a particular construction worker category are paid the same wage, that wage is viewed as the prevailing wage. If there is no clear majority wage, a weighted average of the submitted wages is calculated to determine the prevailing standard.

In the other 39 counties, the prevailing wage determined by the Federal Government is accepted for use in implementing Kentucky's prevailing wage law. These wages are set by the United States Department of Labor on the basis of information collected from contractors and construction worker labor unions. The Federal Government uses a similar rule to that used at the state level to make wage determinations.

Prevailing Versus Actual Wages

We turn now to the relationship between actual and prevailing wages. In this regard, the Kentucky Legislative Research Commission's findings are unambiguous. They state in their Executive Summary, "Currently, neither prevailing wages set by the Kentucky Labor Cabinet or the United States Department of Labor yield prevailing wages that are representative of local wages" (p. ix). Rather, the bias is in the direction of prevailing wages being higher than actual wages.

The source of the discrepancy between prevailing and actual wages is found in the dominance of union scale rates in the data employed in determining prevailing wages. Among wage standards set at the state level, 81 percent of the wage rate data collected at local hearings originates with union workers, despite union workers accounting for only 18 percent of the construction labor force. This leads to union wage scales dominating the determination of prevailing wage standards. The Commission reports that in 372 determinations where there was sufficient evidence to assess union influence, 256 (or 68%) of the prevailing wage determinations were identical to the union wage.

Similarly, the Commission's report presents evidence which indicates that the federal prevailing wage rates used in Kentucky are biased upward. Prevailing wage rates are compared with Bureau of Labor Statistics information detailing actual wage rates in 27 separate instances. The median percentage differential between the prevailing and actual wage rates is 17 percent, in favor of the prevailing wage.

Impact on Construction Costs

Having established that prevailing wages in Kentucky are greater than actual average wages paid, the authors of the report turn to the question of the impact of prevailing wages on construction costs for public projects. They begin their consideration with the following remarks:

Public projects in Kentucky are typically awarded to the lowest bidder. In the absence of prevailing wage laws, contractors are free to select among various mixes of inputs, such as labor and equipment, in an attempt to develop a competitive bid. Prevailing wage laws constrain contractors from one avenue by which they can reduce bids and, therefore, the costs of construction. To the extent that prevailing wages are higher than the wage that would be paid to at least some workers, paying prevailing wages could increase the costs of construction (p. ii).

Having said that, they add:

When faced with paying higher wages, contractors will attempt to shift from using low skilled labor to more productive labor or increase the use of equipment. This substitution tends to offset some of the additional costs associated with prevailing wages. Although there are several studies which conclude that there are no additional costs associated with prevailing wages, there are a number of technical problems which raise doubt about their validity.

The studies just referred to are numerous, and are summarized and discussed on pp. 45-70 of the report and in an Appendix B.

In dealing with the specific issue of the impact of prevailing wages on construction costs, the fundamental problem is an efficiency-wage consideration. In short, do the higher wages implicit in the prevailing (or union) wage premium lead to increased worker productivity, either through greater work effort, the hiring of more productive workers, or a higher capital-labor ratio. Evidence on this point is provided by Steven Allen, who concludes that an earlier union vs. non-union productivity advantage had disappeared by 1982. If this is the case, productivity feedbacks are a non-issue.

The Kentucky Legislative Research Commission, not content to rely solely on the Allen findings, collected data that describe wages paid to workers on prevailing wage covered projects as well as wages paid to the very same workers when employed on private (non-prevailing wage) jobs. The data set for this comparison consisted of 283 workers (90 on road projects, 47 in state non-road activity, and 146 employed on education projects). The results of the analysis are summarized as follows:

Overall, the 283 workers were paid $86,029 more as a result of the prevailing wage. This indicates that prevailing wages resulted in a 24% increase in the wage costs for these projects ... It is important to understand that this estimate does not imply that prevailing wages increased the cost of these projects by twenty-four percent. Rather, it indicates that the wage portion of construction costs was twenty-four percent higher as a result of the prevailing wage. It is argued by some that the additional wage costs of requiring prevailing wages are offset because contractors substitute more productive workers or use more equipment. Because this analysis only examines the workers that actually worked on the project, it incorporates any productivity gains achieved by this type of substitution. Since prevailing wages paid on the projects are still higher than the wages that these workers are paid on private projects, any productivity gains do not effect the full wage cost associated with the prevailing wage requirements (pp. 57-59).

The total cost effect of the 24 percent prevailing wage premium observed in these data will depend on the relative importance of labor costs. If labor costs represent one-half of total construction costs, the taxpayers' bill for public construction projects will be 12 percent greater. If they are 40 percent of total costs, the extra cost of public construction will be 9.6 percent. A labor cost share of 30 percent will produce a 7.2 percent extra project cost burden.

Some Additional Considerations

Beyond the questions of determining the prevailing wage premium and assessing its impact on public construction costs, the Kentucky report also responds to a variety of arguments made by proponents of prevailing wages. Among them is the proposition advanced in the often-cited study entitled, Losing Ground, that reductions in tax revenues resulting from declines in construction worker earnings exceed any construction cost savings associated with repealing prevailing wage laws.

This argument is, at best, a strange one. On the one hand, it suggests that repeal of prevailing wage laws in nine states reduced the annual earnings of construction workers by an average of $1,835 (1991 prices) while, on the other hand, arguing that the construction cost savings associated with these lower earnings are less than the taxes that would be paid on the lost earnings. For this to be true, each dollar of lost earnings must reduce construction costs by less than the marginal tax rate on income in a state. For example, in a state with a five- percent marginal tax rate, construction cost savings per dollar of lost earnings would have to be less than five cents. Given the findings that the very same workers were paid 24 percent more on prevailing wage projects than when working in a non-prevailing wage setting, this does not seem to be applicable in the case of Kentucky.

There are other problems with this argument. The report quite properly concludes that this analysis does "not provide sufficient evidence to conclude that the repeal of prevailing wages would result in lost tax revenue that would exceed reductions in construction costs" (p. 67).

Additional Evidence

Confirmation of these findings is provided by an exercise in polling conducted by the Kentucky Legislative Research Commission. It surveyed six different groups of Kentuckians involved with prevailing wage projects: contractors, representatives of unions with members working for contractors, city government officials, county government officials, local school district representatives, and municipal utility officials. Of these groups, two expressed strong support for the prevailing wage system - contractors with union workers and union officials. This is exactly what would be expected given the finding that there is a significant prevailing wage premium that arises because prevailing wages tend to be the union wage. Contractors with union workers have committed themselves to paying that premium and, consequently, are at a competitive disadvantage relative to contractors with non-union workers. Therefore, it is in their interest to support government intervention that will mandate higher labor costs for their competitors.

As to the union officials, their support for prevailing wages has an obvious motivation--higher wages for union members and more money for unions. The prevailing wage premium is nothing more than an economic rent that is being captured by union members. The recipients of economic rents generated by a policy regimen are not disinterested observers of that regimen. Quite the contrary. They have a vested interest in supporting the policy involved. Perhaps this explains the near unanimous approval of the prevailing wage system expressed by union officials, 97.7% of whom feel that the effects of prevailing wages are either somewhat positive or very positive.

The remaining groups bear the burden of these economic rents, either directly as the funders of prevailing wage construction projects, or as a construction company with non-union workers that has a competitive disadvantage foisted upon it by the government regulatory apparatus. Among this latter group, when polled as to the effects of prevailing wages, nearly two-thirds replied that they were somewhat negative or very negative. By contrast, three-fourths of contractors with union workers feel that the prevailing wage system has either somewhat positive or very positive effects.

Among those involved with the financing of prevailing wage projects, attitudes toward the system are decidedly negative. When asked whether prevailing wages increased construction costs, about 90 percent of city, county, school district, and municipal utility officials replied in the affirmative. Also, when polled about the impact of the prevailing wage system on the quality of construction, only a handful of officials (less than 10 percent) expressed the view that it improved quality.

Collectively, the results of the opinion polls administered by the Kentucky Legislative Research Commission reveal a set of attitudes that is powerfully consistent with the Commission's conclusion that the prevailing wage premium is directly associated with the union wage premium in the construction industry.

Evidence from Nearby States

An interesting question is whether Kentucky is somehow unique in its prevailing wage environment. To explore this issue, conditions in two nearby states are worth noting. First, there is the case of Ohio, where a five-year exemption from its prevailing wage law is running its course. The Kentucky report summarizes a preliminary analysis of the Ohio experience as follows:

The Ohio Legislative Budget Office is currently studying the effects of a five-year exemption of the state's prevailing wage law. Contractors that were awarded construction projects during the exemption were asked to estimate their bid amount if prevailing wages had been applied. Based on this comparison, interim reports on the study conclude that prevailing wages increased the costs by 10.5% on projects where a savings could be estimated (p. 53).

A similar series of events occurred in Michigan. In this case, a federal district court ruling (December 1994) held that Michigan's prevailing wage law was in conflict with a federal pension law, ERISA, and that ERISA takes precedence over the state legislation. The decision was later reversed by an appellate court (June 1997), but in the interval the Michigan prevailing wage law was not enforced. This affords the opportunity to assess the cost and employment effects of a specific minimum wage law during a period when it is not operative. An analysis along these lines is quite revealing. It shows that during the time when the Michigan prevailing wage law was in abeyance, employment in the construction industry surged--even after adjusting for variation produced by seasonal, weather, and business cycle conditions.

Perhaps the most revealing statistic in this regard is the ratio of construction jobs created to total new jobs. In an era when the prevailing wage law was enforced (June 1992 through December 1994), 78.61 new construction jobs emerged for every 1,000 new jobs of all kinds. However, from December 1994 through June 1997, a period without prevailing wages, that ratio was 48 percent greater, standing at 116.27.

In addition, the Michigan experience confirms the existence of a prevailing wage premium. The median percentage differential between the prevailing wage and the market wage in nine different construction occupations in 1992 was 31.6 percent in Wayne County, and 39.3 percent in Monroe County. In both cases, the median differential occupation was that of carpenter. These wage premiums, for a period when the prevailing wage law was in force, translate into higher construction costs. By contrast, when 20 different projects were examined for the period characterized by no enforcement of the prevailing wage law, "On average ... the savings were well above ... 10 percent" (Vedder, p. 14).

In summary, the experiences of two nearby states, Ohio and Michigan, are quite consistent with the findings reported by the Kentucky Legislative Research Commission. This indicates that the Kentucky analysis is not an anomaly that can be disregarded in the broader perspective of assessing the effects of prevailing wages.

Concluding Remarks

The report of the Kentucky Legislative Research Commission on prevailing wages provides a rather thorough indictment of the policy of mandating such wages for publicly financed construction projects. As such, it offers an excellent summary of the case against prevailing wages and their tendency to restrict people from operating in a free market, where they would otherwise be able to allocate resources and use production factors most efficiently. Thus, prevailing wages retard job creation and lead to slower economic growth.

The Kentucky report should be studied by anyone interested in the issues that attend the prevailing wage question.

* Lowell Gallaway and Richard Vedder are both distinguished Professors of Economics at Ohio University, and are the authors of numerous books and articles on labor economics.




1. An Analysis of Kentucky's Prevailing Wage Laws and Procedures, Staff Report, Kentucky Legislative Research Commission (Lexington, Kentucky, December 2001).

2. "Can Union Labor Ever Cost Less," Industrial and Labor Relations Review, April 1988, pp. 347-73.

3. Peter Phillips, Garth Mangum, Norm Waitzman, and Anne Yeagle, Losing Ground: Lessons from the Repeal of Nine Little Davis-Bacon Acts(February 1995).

4. Ohio Legislative Budget Office, A Study of the Effects of the Exemption of School Construction and Renovation Projects from Ohio's Prevailing Wage Law, An Interim Report of a Five-Year Study - Year Two (January 2000).

5. For a description and analysis of this situation, see Richard Vedder, Michigan's Prevailing Wage Law and Its Effects on Government Spending and Construction Employment (Midland, Michigan: Mackinac Center for Public Policy, 2001).

The 1990 West Virginia Prevailing Wage Law Study UWV

Back in the United States, opponents of the federal and state Davis-Bacon acts charge that the statutes inflate the costs of government construction projects and should be repealed. However; there seems to be a paucity of available studies providing the facts and figures to support such contentions. For that reason, the Review chose to publish The 1990 West Virginia Prevailing Wage Report.

The report was prepared by the School of Business and Management at the University of West Virginia College of Graduate Studies in Institute at the behest of a state delegate. It was felt that a study by an objective academic body would avoid the charge of bias that labor or management would level, had one or the other commissioned the work. The study found that the state's prevailing wage law raised the average cost of public construction by about thirty percent.

An Evaluation of the Impact of the Davis-Bacon Act Berg, John and Erickson, Ralph

To prevent the exploitation of local labor markets by itinerant contractors during the Great Depression, in 1931 the United States Congress passed the Davis-Bacon Act. This legislation requires contractors who bid on federally financed building projects to pay construction workers the locally prevailing rates in wages and benefits.

In An Evaluation of the Impact of the Davis-Bacon Act, government economists John T. Berg and Ralph C. Erickson review the historic impetus which gave birth to the Act, offer a critical analysis of the arguments propounded by its modern apologists and examine the reasons behind the growing support for its reform or outright repeal. The authors found that forced compliance with the Act annually raises public construction costs between $1 to $2 billion and of the federal-aid highway program, of which they are most familiar, $284 million in 1984.

Public sector unions favor retention of the Act since it has an inflationary effect upon their members' wages. Even though the last year has witnessed repeal of "little Davis-Bacon Acts" in Arizona, Idaho, New Hampshire and Colorado, unions continue to lobby against its repeal on both the state and federal levels. During the last session of Congress, they successfully opposed efforts on Capitol Hill to reduce defense expenditures by exempting contractors from the prevailing wage statute on military projects of less than $1 million.

The Service Contract Act of 1965: Time to Revise or Repeal Burns, Beverly

Effusing with altruistic intentions, in 1965 the United States Congress passed the Service Contract Act, ostensibly to assure that service employees were provided with minimum wages and benefits. It was felt that this lowest scale of workers, used by private contractors on federal projects, had slipped through the coverage extended to other employees by the 1931 Davis-Bacon Act and the 1936 Walsh-Healy Public Contracts Act. The question being asked today is how well has the Service Contract Act fulfilled its purpose?

Almost from the day of its enactment, problems concerning wage determinations, the rapid turnover and rebidding of contracts, and a precise definition of who constitutes a service worker led to amendments of the Act in 1972 and 1976. In The Service Contract Act of 1965: Time to Revise or Repeal, Beverly Hall Burns provides a legalistic analysis of the Act, its history, the deficiencies that plague it and court cases which have sprung from it. After a thorough evaluation of a piece of legislation still avidly supported by the AFL-CIO, the author concludes that the Act "must either be repealed... or completely overhauled to deal with the multitude of problems it has created."

Prevailing Wage Laws of the States Thieblot, Armand

Congress passed the Davis-Bacon Act in 1931, then designed to counteract the effects on local contractors of cutthroat competition by itinerant contractors scouting the country during the Depression and underbidding local firms on government projects. The act requires contractors to pay "prevailing wages" on construction projects funded with federal money.

The passage of the Davis-Bacon Act was followed by a flurry of similar laws being enacted in many states, so-called "little Davis-Bacon Acts," which impose wage requirements on construction projects funded with state, and sometimes local, money. Thirty-seven states to date have such laws.

The Davis-Bacon Act also has stirred a considerable amount of controversy which has increased ever since its inception. Nearly every study conducted on the act has concluded that it has an adverse effect both in terms of economic cost and employment opportunities of nonunion and minority employees. Under the administration of the Department of Labor, critics point out, the effect has been to favor large unionized urban contractors at the expense of open shop and small local contractors. Cost estimates have pegged the extra expense resulting from the act at as high as 34 percent over comparable private industry construction, primarily due to the Labor Department's method of calculating "prevailing wages," which more often than not leads to the adoption of union scale rates. Not surprisingly, Davis-Bacon finds its only supporter in organized labor and its friends.

Academic and professional research and cost analyses concerning Davis-Bacon have generally been sporadic, fragmented and usually oriented toward the federal act. This issue of Review features a comprehensive summary by Armand J. Thieblot, associate professor, University of Maryland, of the state-level prevailing wages laws. A comparison reveals wide variations among state laws, according to Dr. Thieblot, and a historical examination shows that some laws even pre-date the federal act, while others were passed relatively recently. The general trend has been either toward repeal or limiting coverage, but some states however, have decided to expand.

Dr. Thieblot's work is a welcome addition to the body of knowledge of Davis-Bacon and its state level counterparts. In addition to being a guide to the national picture of prevailing wage laws, it provides detailed analyses of each one of the state acts, historical understanding of their development and a sense of future direction. This is part of a larger study of federal and state prevailing wage legislation being completed by the Industrial Research Unit of The Wharton School, University of Pennsylvania.