House Bill 4329 as introduced
House Bill 4383 (Substitute H-1)
House Bill 5188 as introduced
Sponsor: Rep. Robert Gosselin
First Analysis (12-5-01)
Committee: Employment Relations, Training and Safety
Peter Philips, a professor of economics at the University of Utah, published a study in February 2001 called "Comparison of Public School Construction Costs in Three Midwestern States That Have Changed Their Prevailing Wage Laws in the 1990s." See BACKGROUND INFORMATION below. Philips notes that "proponents and critics of prevailing wage regulations have debated the merits of the regulations for some time. Proponents argue that the regulations promote the development of a skilled labor force in construction, improve work place safety, encourage quality construction, increase apprenticeship training, and provide career opportunities in construction for local citizens. Proponents emphasize that prevailing wage regulations also induce contractors to provide health insurance and pension coverage that otherwise would be absent. In contrast, critics of prevailing wage regulations concede some of the foregoing position and contest others. But the main argument of critics of prevailing wage regulations is the contention that these laws raise public construction costs: they assert that when regulations are applied they raise construction costs, and when they are eliminated, public construction costs go down. Generally they claim the magnitude of savings ranges from 10 to 30 percent."
In Michigan, the prevailing wage law has been repealed and subsequently reinstated during the past decade, following a court challenge. Throughout, the enforcement of the act has been the responsibility of the Wage and Hour Division of the Department of Consumer and Industry Services. See BACKGROUND INFORMATION below.
In addition to the argument that prevailing wage regulations increase construction costs when they are in effect, some policy makers who favor modifications to, or outright repeal of, the statute argue that prevailing wage laws are an example of unacceptable government interference in labor markets. These free market advocates say that neither government agencies nor trade unions should intervene in labor markets, either to regulate or to influence wages, and they hold that any intervention to establish wage rates is bound to create artificial wage limits, as well as to promote a proliferation of job categories that deter needed flexibility and efficiency when managers deploy workers at construction sites.
Recently, those who have sought to modify (rather than to repeal) the statute have proposed that school officials, whose budgets are constrained during periodic economic down-turns, especially need relief from prevailing wage rates when they bid school construction projects. To that end, legislation has been introduced.
THE CONTENT OF THE BILLS:
Generally, Michigan's prevailing wage law, Public Act 166 of 1965, requires that wages and fringe benefits paid by contractors on state construction projects financed in whole or in part by state funds, including local school construction, must not be less than the wages and fringe benefits prevailing in the locality where the work is to be performed. The law is administered by the Department of Consumer and Industry Services, which determines the required prevailing rates in various areas of the state.
House Bill 4383 would amend the prevailing wage law (MCL 408.551) to exempt from the law those construction projects involving public schools. Local and intermediate school districts and public school academies would be excluded from the definition of "contracting agent" for purposes of applying the prevailing wage requirement. Further, House Bill 4329 would amend the Revised School Code (MCL 380.503 and 380.1311e) to delete requirements that public school academies and strict discipline academies adhere to the requirements of the prevailing wage law.
In addition, the Revised School Code requires that the board of a school district or the board of directors of a public school academy obtain competitive bids on all material and labor for new school buildings and additions to, repairs of, and renovations of existing school buildings. House Bill 5188 would amend the Revised School Code (MCL 380.1267) to specify that a board or board of directors could not "discriminate against a bidder because of religion, race, color, national origin, or sex, or because of the status of its employees with respect to collective bargaining representation."
Court decisions to invalidate and then reinstate the prevailing wage law. Between November 1994 and June 1997-a period of about 21/2 years-Michigan operated without a prevailing wage law. The state prevailing wage act was rescinded by the U.S. District Court for the Eastern District of Michigan on November 21, 1994 when that court ruled that certain administrative provisions of the act were pre-empted by the federal Employee Retirement Income Security Act (ERISA). The court ruled the prevailing wage law improperly enforced reporting and disclosure requirements, established rules for the calculation of benefits, and imposed remedies for alleged misconduct, all of which were pre-empted by ERISA. Further, the court ruled that ERISA pre-empted provisions concerning an excess benefits cap, and apprenticeship requirements. The court held that the preempted sections could not be severed from the non-preempted portions, and invalidated the entire prevailing wage law which had been in effect since 1965.
However, on June 27, 1997, the U.S. Court of Appeals for the Sixth Circuit overturned the district court decision, ruling that Michigan's prevailing wage law, like other states' prevailing wage laws, were not pre-empted by ERISA, and noting that "[F]ederal statutes will not be held to have pre-empted laws in areas of traditional state regulation unless that is the clear and manifest intent of Congress." [Associated Builders and Contractors, Saginaw Valley Area Chapter v Lowell W. Perry]
Department of Consumer and Industry Services, Wage and Hour Division. The Michigan Department of Consumer and Industry Services determines prevailing rates under the prevailing wage law, Public Act 166 of 1965. The purposes of establishing prevailing rates is to provide rates of pay for workers on construction projects for which the state or a school district is the contracting agent, and which is financed or financially supported by the state. By law, prevailing rates are compiled from the rates contained in collectively bargained agreements which cover the locations of the state projects. The department establishes the prevailing rate for each classification of construction mechanics requested by a contracting agent prior to contracts being let out for bid on a state project. [If a contract is not awarded or construction does not start within 90 days of the date of issuance of rates, a re-determination of rates is requested.] The prevailing wage law requires every contractor and subcontractor to post a copy of the prevailing rates at the construction site. An accurate record showing the name, occupation, and the wages and benefits paid to each construction mechanic must be kept by the employer, and made available to the department for inspection, upon request. A worker who has not been paid the prevailing rate on a state project may file a complaint with the Wage and Hour Division. (The address is: Michigan Department of Consumer and Industry Services, Bureau of Safety and Regulation. Wage and Hour Division, 7150 Harris Drive, Post Office Box 30476, Lansing, Michigan 48909-7976; Telephone: (517) 322-1825.) The department will investigate and attempt to resolve the complaint informally. If voluntary resolution is not successful, the complainant is referred to the prosecuting attorney for enforcement action. The division does not classify workers or resolve classification disputes. For more information, visit the web site at www.commerce.state.mi.us/bsr/divisions/wh.
A comparison of public school construction costs. Peter Philips, a professor of economics at the University of Utah, published a study in February 2001 called "Comparison of Public School Construction Costs in Three Midwestern States That Have Changed Their Prevailing Wage Laws in the 1990s." Philips' study looks at the construction costs of 391 new public schools built between 1991 and September 2000. About half the projects were built with prevailing wage rates (49 percent), while 51 percent were built without them. Thirty-eight percent (38 percent) of the schools were built in Michigan, while Ohio accounted for 36 percent, and Kentucky for 26 percent. Philips selected these three states because during the period of study, Ohio repealed its prevailing wage regulations, Kentucky instated prevailing wage rates for the first time, and Michigan both repealed and then reinstated its prevailing wage regulations (as a result of the court test, above). Consequently, the history of prevailing wage policy in the three states allowed the economist to track wage effects when new schools were constructed in the different regions. He designs an experiment in which he controls for other factors, in order to ascertain whether changes in prevailing wage regulations made a difference in school construction costs.
The study reveals that the average (or mean real, that is to say inflation adjusted) square foot cost of rural schools built without prevailing wages was $96 per square foot, while the square foot cost of rural schools built with prevailing wages was $98 per square foot. Statistically speaking, a t-test applied to the data indicates there is no difference between the two. [A lack of statistical significance means that one cannot be sure there really is any difference in the totals, generally because of their similarity in light of other relevant facts.] The study also reveals that the average square foot cost of urban schools built without prevailing wages was $114 per square foot, while the square foot cost with prevailing wages was $114.34. Again, statistically, the averages are equivalent.
An economist, Philips then arrays the data in a manner that is customary within his professional discipline as a researching social scientist. He uses a statistical model of new public school construction costs, called an ordinary least squares linear regression model. He explains that this model is very commonly used by economists, epidemiologists, and others studying social phenomena. To implement the model, Philips uses the 391 new public schools built in Kentucky, Ohio, and Michigan over the 1991 to 2000 period as data to help predict the effect of the presence of a prevailing wage law (the focus variable) while controlling for nine other variables in new construction. [The variables of the study are total square feet of the project, time (in years), where built (Ohio, Kentucky, or Michigan), whether urban or rural, whether construction was begun in fall, winter, spring, or summer quarter, and whether the school was built with prevailing wage rates.]
When the statistical tests comprising an ordinary least squares linear regression were run electronically against the data, the results indicated that after adjusting for inflation, new public school construction costs in the three states have been rising at 2.9 percent per year from 1991 to 2000. This result is statistically significant, and the reason costs have been rising faster than inflation is because the economic boom has led to a very vigorous boom in building, which in turn has led to heavy demand for construction services. The model also indicates that new school construction costs are highest in Michigan among the three states. New Ohio schools cost 12.6 percent less than Michigan schools, while new Kentucky public schools cost 14.6 percent less than new Michigan public schools. The model also indicates that starting construction in the spring quarter rather than the fall saves 10.9 percent (controlling for all other factors), and these results are statistically significant (which means a policy maker can be confident that breaking ground in the spring will lead to lower cost new school construction than would a fall start).
Overall, the Philips model estimates that controlling for all variables (seasonality, differences in rural and urban construction costs, size of the project, and the state in which the project was built), the use of prevailing wage regulations raised school construction costs by seven-tenths of one percent, an amount that is not statistically significant. In effect, the model says there is no effect on total costs associated with prevailing wages.
In the conclusion of the study, the economist notes that it may be that with more observations, the 0.7 percent higher costs associated with prevailing wage regulations would turn out to be statistically significant. He cautions, however, that the model is based on start costs-accepted bid price. The ultimate cost of a new school includes cost over-runs, and the downstream cost of maintenance.
The House Fiscal Agency notes that House Bills 4329 and 4382 (H-1) could reduce state and local costs in two ways. First, state costs incurred by the Department of Consumer and Industry Services for administering and enforcing the prevailing wage statute would be reduced by an indeterminate amount. Second, to the extent that the exemptions reduce wages paid by contractors that would no longer be covered by the law, local costs for school construction, repair, or improvement projects could also fall.
On the revenue side, the agency notes that state income tax revenue and local income tax revenue, where applicable, could decrease by a very slight amount. This would depend upon the extent to which the bills' provisions lead to lower wages paid by contractors exempted by the bills.
Finally, the House Fiscal Agency notes that House Bill 5188 should have no fiscal impact on the state or on local units of government. (12-4-01)
According to the National Federation of Business Michigan Chapter, elimination of the prevailing wage rate regulation in school construction projects saves taxpayers money. They report that since the repeal of the prevailing wage on school construction projects in Ohio, that state's portion of construction costs appropriated through the Ohio School Facilities Commission has been $2.68 billion since 1998. The average savings as a result of prevailing wage repeal have been between five percent and 10.5 percent. Overall, they estimate that taxpayers in Ohio have saved between $135 and $270 million.
Further, they point out there is solid and recent evidence that Michigan's prevailing wage rates increase construction costs in this state. After Michigan's prevailing wage rate was reinstated in June 1997 following a 21/2-year court-imposed hiatus, costs on construction projects increased between six and 15 percent, using the same contractors.
Proponents note that eliminating prevailing wage rates is timely, given the state's budget shortfall, and the recent cutbacks in school funding. They say that one way to increase school districts' spending power in this period when tax dollars are in short supply is to eliminate the need for school board officials to pay higher prevailing wage rates when they build new schools. They argue that Michigan's prevailing wage law is outdated and wasteful, and that it should not be used on school construction projects.
Proponents of the bills refute five claims proffered by those who support prevailing wage rate regulation. First, to those who claim that prevailing wage laws protect worker safety, they respond that Michigan's prevailing wage law is completely silent on issues of safety. OSHA and MIOSHA regulate safety at work-sites, and the same regulations apply to all work-sites, whether or not prevailing wage rates are in effect. Further, in Michigan an analysis of accident reports from 1996 to 1999 showed that unionized employers had three times as many willful MIOSHA violations as non-union employers. The same analysis showed no significant correlation between injury rates and the presence or absence of the state prevailing wage rate. ("Prevailing Wage Rates as They Relate to Accident Rates", memo authored by Kalmin Smith, Deputy Director, Michigan Department of Consumer and Industry Services, 7-11-00)
Second, to those who claim the prevailing wage law guarantees construction quality, proponents of the bills note that prevailing wage projects are subject to the same building codes as any other project, and that building inspectors use the same standards for all.
Third, to those who claim that repealing the prevailing wage law would reduce worker wages, proponents of the bills note that according to the Bureau of Labor Statistics, in the two years prior to Ohio's repeal of a prevailing wage on school projects, construction workers' wages increased 4.5 percent, while in the two years following the repeal, wages increased 6.9 percent.
Fourth, to those who claim the prevailing wage law allows better training for workers, they note that Michigan's prevailing wage law does not require training of any kind. Oversight of training programs falls to the Bureau of Apprenticeship Training in the Department of Labor, and that governmental agency approves both non-union and union-sponsored training programs.
Fifth, to those who claim that the prevailing wage law ensures a local and diverse workforce, proponents of the bills point out that the state prevailing wage law does not require that workers be local, or for that matter, that they have citizenship status. Indeed, many "local" contractors are merit shot (or non-union) contractors who run small businesses and who cannot afford to meet the excessive regulatory demands of a prevailing wage job-site. In addition and according to the Bureau of Labor Statistics, more than 80 percent of all women (88.5 percent), African-Americans (83 percent), and Hispanics (89 percent) choose non-union employment opportunities.
According to committee testimony, prevailing wage rates are based on skilled construction trade jurisdictions, job classifications, and the geographic jurisdictions of construction unions. Due to the complex maze of craft jurisdictions, it is difficult to monitor prevailing wage rates, a responsibility that falls to the Wage and Hour Division of the Department of Consumer and Industry Services (CIS). Those familiar with enforcement report that there are no written job descriptions to distinguish one classification of construction worker from another, and each has entirely different wage and fringe benefits rates. Instead, unwritten rules developed by different trade unions are applied, often changing from job to job. Consequently, the CIS must often resolve work classification disputes by relying on the interpretations of union officials. These circumstances are unworkable for non-union contractors who do not organize their workforce along artificial and inflexible job classifications that are found in union collective bargaining agreements, but instead train workers to carry out a number of skilled tasks and to shift between them at a job site. It is not sensible to have a prevailing wage law that forces non-union employers to change their more efficient mode of operation whenever they work on state-funded or public school projects.
Opponents of these bills say they offer school boards the chance to become penny wise and pound foolish. They observe that the final cost of new public school construction includes cost overruns and downstream maintenance costs. They argue that the higher wage rates required by prevailing wage regulations are likely much exaggerated. Nonetheless, the presence of the wage regulations ensures that all contractors bidding on the job will use skilled labor when building the school. In this way the prevailing wage regulations offer school board officials some assurance that the project will be skillfully built, and that workers on the job will be carefully managed. Consequently, prevailing wage regulations provide some assurance against cost overruns and downstream maintenance costs.
Opponents note that prevailing wage contracts provide the written framework within which management and labor can dedicate pennies per worker on each project, to job training and to pensions. Absent the written agreement, there is no guarantee to workers that employers will dedicate some of their profit to these ends. Indeed, pension benefits for non-union workers are rare, according to committee testimony. Absent the written agreement, there is no guarantee that newcomers to the construction trades will learn productive and safe work habits and skills, since their opportunities for union-sponsored apprenticeships and training will undoubtedly diminish.
According to a study conducted by the economist Peter Philips-a study designed to learn the effects of prevailing wage rates on new school construction during a nine year period in Michigan, Ohio, and Kentucky--there is no effect on total costs associated with prevailing wages.
The study findings indicate that while costs increase seven-tenths of one percent (0.7 percent), that finding is not reliable, or statistically significant.
Published in February 2001, the Philips study concludes that a simple comparison of the mean (or average) inflation-adjusted square foot cost of building 391 new public schools broken down by urban and rural schools finds no statistically significant difference between those public schools built with prevailing wages and those public schools built without them. Further, a more complex statistical model that estimates new public school construction costs based on the size of the project, whether it was an urban or rural school, which state built the school, and at what time of the year the school was built, again finds no statistically significant effect on total new school construction costs associated with whether or not the school was built with prevailing wages. While no effect of prevailing wage regulations is apparent, the study did reveal that school boards can save 10 percent on new school construction costs by starting in the spring, and not breaking ground in the face of winter weather.
The design of the Philips study was questioned during committee deliberations. Proponents of the legislation argue Philips asked the wrong question. He should not have asked: Do prevailing wage rates affect construction costs? This, they argue, is the wrong research question; instead he should have asked: Do prevailing wage rates affect labor costs? Further, proponents of the bills argue that Philips' use of least-squares multiple regression analysis to model the data is a flawed method of analysis. They say the multiple-sample nature of his dataset requires other kinds of statistical tests.
The Philips experiment design is entirely appropriate given his dataset, and his statistical research methods are fully in keeping with the theory and practice of statistical analysis within his discipline (and other social sciences). To challenge a researcher's question is always an acceptable rhetorical device, and can serve to clarify or sometimes confuse the nature of the policy problem. It cannot, however, discredit the findings of that researcher's experiment.
The following groups have testified or provided written testimony in support of the bills: The Hillsdale Policy Group; the Iosco Regional Educational Service Agency; the Michigan Association of School Boards; the Michigan Grocers Association; Oakland Schools; the National Federation of Independent Business-Michigan; the Grand Rapids Chamber of Commerce; the East Grand Rapids Public Schools; and, the Kent Intermediate School District.
The Michigan Road Builders Association is neutral on House Bill 4383 (H-1) and has taken no position on the other bills. (12-4-01)
The following groups have testified or provided written testimony in opposition to the bills: the Michigan Chapter of the Associated General Contractors; the Michigan State AFL-CIO; the United Auto Workers International; the Michigan Federation of Teachers and School-Related Personnel; the Michigan Education Association; the National Electrical Contractors Association; the Mid-Michigan Mechanical Contractors Association; the Michigan State Building Trades; the IBEW Local 58; and, the Service Employees International Union (SEIU).
This analysis was prepared by nonpartisan House staff for use by House members in their deliberations, and does not constitute an official statement of legislative intent.