The Failure of Arguments Supporting Prevailing Wage Laws and a New Evaluation of the Benefits of Repeal by A. J. THIEBLOT, Ph.D.*

A much-reported recent study claims that repealing the Davis-Bacon Act (the federal prevailing wage law for construction of public works) would cost more in lost taxes than could be saved in lower construction expenditures, and would result in more construction injuries and deaths. Those claims are unsupported. The facts support savings in excess of $1.5 billion annually to the federal government, and possibly fewer construction injuries, by repealing Davis-Bacon. States with prevailing wage laws would also realize significant savings from repeal. Aggregate savings from eliminating all prevailing wage laws could exceed $4 billion a year.

I. Introduction

The Davis-Bacon Act is the federal prevailing wage law for the construction industry. It requires contractors on public works involving the federal government to pay their workmen at least the wages and fringe benefits found by the Department of Labor (DOL) to be "prevailing" for similar work in the same locale. Although the Act was created to cure problems that predated the Great Depression, it has remained on the books in largely unmodified form since 1931.1 Most states passed similar laws for public works at various times, and by 1975 prevailing wage laws existed in 41 states.

In the late 1970s, it began to dawn on some policy makers that the problems prevailing wage laws were designed to correct no longer existed, and that the laws forced governments to spend more money to get the same quality of construction that private buyers could routinely get for substantially less. Thus, a movement to eliminate or modify the laws began in legislatures and courts, with some success. Since 1979, nine states have repealed their prevailing wage laws and the laws of two other states were declared unconstitutional.

The momentum building in the repeal movement was shaken in early 1995, however, by a new study (actually a "working paper" released over the Internet) of prevailing wages which seemed to contradict conventional wisdom about the impact of repeal.2 The study, titled Losing Ground: Lessons from the Repeal of Nine "Little Davis-Bacon" Acts, was authored by Peter Philips, Garth Mangum, Norm Waitzman, and Anne Yeagle, all of the University of Utah. It used multiple regression analysis and other sophisticated statistical tools to claim a negative economic and social impact of repealing prevailing wage laws at both federal and state levels. Among its findings were that governments would lose up to $1.5 billion more in lost tax revenue than they could save through lower construction costs, and that repeal would cause increased construction injuries and deaths, resulting in billions of dollars of associated costs.

The Utah study contradicted all previous researchers of this topic and invited further analysis—especially since its conclusions were being used to influence public policy. Within weeks of its appearance, it was being quoted in the U.S. Senate and House of Representatives, in state legislatures in Pennsylvania, Michigan, Indiana and Ohio, and elsewhere. The study does not deserve such consideration because its methodology and its findings—for all their apparent sophistication—are deeply flawed. Exposing the errors of the study is a tedious undertaking, but not entirely negative. One statistical technique developed by the Utah study's authors has actually given us a new approach to estimating the real economic impact of prevailing law repeal and confirms that such repeal would, indeed, provide the billions of dollars of savings that were previously estimated by other means.

II. Synopsis of Major Conclusions of Utah Study

The Utah study derives its conclusions from massive regression analyses (using numbers from more than 27,775 data points—bits of information), anecdotal evidence, economic reasoning, and examination and manipulation of time series data on employment and earnings. The study's statistics are derived from the repeal of the Utah prevailing wage law in 1981, or from repeals in the other eight states that eliminated prevailing wage laws from 1979 to 1987. These are projected to the national level to anticipate the impact of repealing the federal Davis-Bacon Act.

The study's three major economic claims are summarized below:



  1. Repeal drives wages down. Aggressive competitive patterns brought on by the 1981 repeal in Utah drove wages down below the market rate for all construction workers, not just those on state-financed jobs. Average earnings fell $1,835 per worker in real terms (as denominated in 1991 dollars) in Utah and the eight other states where prevailing wage laws were repealed.*
  2. Government does not break even. For Utah's 1991 employment figures and state tax rates, decreased construction worker earnings resulted in lower tax revenues to the state of $6.8 million. On the federal level, repeal of Davis-Bacon would lower federal income tax collection by $1 billion to $2 billion in real terms, more than what could be made up through lower building costs.
  3. Repeal causes injuries. Repeal of prevailing wage laws results in additional injuries to construction workers in repeal states. On a national basis, this amounts to 76,000 additional workplace injuries per year (and 130–150 additional deaths), and leads to greater workers' compensation costs totaling $3 billion per year, of which $300 million would be passed on to the federal government as increased costs on public works.

* This Utah study claim of reduced earnings is actually more than a little elusive. It is presented in its executive summary as a "real" earnings loss of $1,477 as denominated in 1994 dollars (i.e., conformed to the purchasing power of 1994 dollars by inflating earlier dollars by the amount of change in the Consumer Price Index). An alternate number, also said to be denominated in "real" 1994 dollars, is found on page 16, where the cost to the same construction workers in the same states is said to be $2,016, or 36 percent more. This deconstructs to $1,835 of "real" earnings loss as denominated in 1991 dollars. No explanation is provided for the difference. On page 23, in the regression analysis, the loss is said to be $1,350 in "real" 1991 dollars. Finally, in another summary on page 68, it is alternatively $2,169 in 1991 dollars (the difference between $24,317 before repeal and $22,148 after repeal), which is roughly $2,386 in 1994 dollars, or $1,477 in 1994 dollars, depending on the analysis used to determine it.

There are many other assertions in the Utah study. Repeals are said to lead to a less skilled construction work force, lower work force productivity and increased cost overruns; and a shift towards newer, smaller, less competent contractors, greatly decreased construction training and less minority representation in training programs; and finally, a rise in the use of expensive change orders.*

* The study's claim that the repeal also exposed government purchasers of construction services to overrunning costs was based on several charts to the effect that cost overruns on highway and heavy jobs in Utah increased after repeal. Since allowing cost overruns and change orders is a matter entirely under the control of the contracting authority and is unrelated to prevailing wages, this part of the Utah study was not further analyzed. The study also claimed that repeal had intensely negative impacts on apprentice training and minority development in the industry. Although these matters are important, they are secondary and proper analysis of them would unduly burden this paper. Nevertheless, I note in passing that my own study of minority employment in various states found an almost perfectly inverse relationship between the proportion of black workers and the strength of a state's prevailing wage law, even when controlled for minority population proportion. That is, even in states with similar proportions of blacks in populations, those with stronger prevailing wage laws almost always saw fewer blacks working in construction.

In sum, the Utah study maintains repeal of prevailing wage laws was or would be an unmitigated disaster, negatively affecting construction companies, union and nonunion construction workers, taxpayers, and government purchasers. Any savings from repeal would be exceeded by tax losses even if they were not offset by cost overruns. The country would suffer from having to make up the tax losses, from absorbing large new costs associated with workers' compensation claims, from an increased accidental death rate, etc. Overall, the Utah study concludes that at certain anticipated savings and cost rates repeal of Davis-Bacon might add $4 billion to $5 billion per year to the net cost of government construction.

Are the Utah study's authors right that society will become unglued simply because government ceases to require contractors to pay their workers above-market wages? Are they right that prevailing wage laws save money? A closer look at their analysis shows that the actual facts support the opposite conclusions. It is repeal rather than retention of prevailing wage laws that will save the government substantial amounts of money. Repeal will also free the construction industry to modernize its methods of training and labor deployment. What follows next is an analysis of the Utah study's three main claims, showing where they are wrong.

III. Examination of and Responses to Major Conclusions of the Utah Study

Utah Study Claim #1: Repeal Drives Wages Down

The Utah study asserts that repeal of the state's prevailing wage law in 1981 drove construction wages down for both union and nonunion construction workers in Utah. The falling wages are explained by a three step hypothesis:



  1. Repeal reduces the competitive edge of union contractors who are often larger and more experienced, thereby decreasing the number of such contractors and the number of their union employees;
  2. Smaller, weaker contractors spring up to replace the existing union firms and compete for government work; and
  3. The increased competitive pressures produce intense competition and an "overheated bidding process" in which "low-balling contractors" "bid jobs down" across the board, thereby lowering wage rates below the market level.

In this way, the study says, Utah construction worker earnings, which averaged about 125 percent of average non-agricultural earnings in the twelve years before repeal, fell to 103 percent by 1993, twelve years after repeal.

This analysis is almost totally incorrect. The economic theory is wrong. Supply and demand of anything, including labor, are related by prices set by the combined self interests of buyers and sellers. New demand for an existing supply never causes the clearing price to fall.* Suppose a contractor wins a post-repeal government building project and needs to add to his work crew. If enough workers are available and willing to work, he hires them, paying what they're willing to work for—the definition of the market rate. (He cannot pay less, because they will not work for it, and he need not pay more.) The market-clearing price remains constant as long as there are enough workers willing to work at the market rate. But if the available supply of workers is not enough, a contractor can induce more only by offering rates that are higher than market. In either case, it cannot result in bidding wages down below the market rate. Thus, the Utah study is wrong in theory.**

* In the typical depiction of a demand curve for labor, it is downward sloping to the right, indicating that if the price of labor decreases, the quantity demanded will rise. The slope of this curve is what one of the Utah study's regression analyses set out to discover when estimating the increased employment that might come about as a result of reduced wage rates. But that is not what we are discussing here. What we are looking at here is exogenous demand, which results in the demand curve shifting outward to the right. If an upward sloping supply curve can shift in proportion, the market-clearing price will remain constant; if not, the price will rise. This can be confirmed in any standard economics text.

**A market condition does exist in which wages can be bid down. It is known as monopsony in economic theory, and is the rarely found situation of the single buyer in a competitive marketplace. Some have argued that monopsony characterized construction labor markets during the Depression. Between 1926 and 1933, for example, the volume of new construction in the United States declined from over $12 billion to under $3 billion, and employment in the industry fell from over 800,000 to under 400,000. Earnings also plummeted, from $1,700 a year down to $820. Simultaneously, the proportion of public construction shot up from the traditional 20 percent to over 55 percent. Contractors on public works in that time period could offer below-market wages which workers would have to accept because there were no alternatives. It could be argued that while government construction was about the only game in town, government contractors were agents of a monopsony. Today, federal construction is back down to about 20 percent of the industry's activity, so there are about four times as many alternative jobs at any given time. Furthermore, there are active safety nets in place for the unemployed. No monopsony power lodges in the federal government's 20 percent market share, and therefore there is no need for the likes of a prevailing wage law to protect wage earners from its monopsonistic excesses.

This is theoretical, but what can be learned from the actual results of repeal? After repeal in Utah (a) larger firms were replaced by smaller ones, (b) the number of contractors competing for work decreased rather than increased (so competition decreased rather than increased), and (c) despite the study's repeated assertions to the contrary, average construction wages did not fall, but rose.

Figure 1 plots the number of contractors in Utah in the four years prior to the 1981 prevailing wage law repeal, and in the four years thereafter. Clearly, the number of contractors fell after repeal, while the number of employees per firm increased—the opposite of what the Utah study asserts. These data show no evidence of bitter competition or lower construction earnings following repeal.

Figure 1
Number and Size of Utah Contractors, 1977-1985


Source: Utah Department of Employment Security

Figure 2 plots the average monthly construction wage rates for the 10 years before and after repeal, showing somewhat slower growth rate in the early 1980s than during the boom years of the late 1970s—but no earnings decrease. For qualitative comparison, the average monthly earnings for employees in the retail trade are also plotted on the graph.3 Notice that the earnings pattern for construction workers follows that of other Utah employees during the same time period, and thus cannot reasonably be attributed to repeal of prevailing wages. During the Carter years of double-digit inflation (and for a few years after), almost all average wage rates in Utah and other states declined in real terms, regardless of prevailing wage rate status.


Figure 2
Average Monthly Earnings, Utah

Source: Utah Department of Employment Security

Since average construction wages did not decrease in Utah during the period of prevailing wage repeal,* the decrease cannot be attributable to increased competition (which in any event did not increase). Therefore, there is no need to invert economic theory to accommodate the effect, since it did not occur.

* The claim that construction wage rates fell is predicated on their being measured in "real" dollars of 1991 or 1994 value, inflated to earlier years by the Consumer Price Index. The number of dollars actually received by the average construction workers increased during the period.

This leaves one final aspect of the Utah study's chain of logic to be analyzed, that is, the assertion that average construction wages fell with respect to average wages in other fields due to prevailing wage repeal. Here, although average construction wages did fall with respect to others, it is highly unlikely that the fall was the result of prevailing wage repeal. Why? Because relative construction wages fell similarly in states that never had prevailing wage laws, and even in all of the states which retained them during this period.

Figure 3 contrasts relative earnings among construction workers against all workers in the state. The Utah study correctly points out that, over the 15-year period from the late 1970s to the early 1990s, average construction earnings in Utah declined with respect to the average for all workers. But taken alone, this is misleading. In the figure, states are grouped into categories, with the taller (back) lineup showing the average ratio of construction earnings to all earnings during the period 1976–1979; and the shorter (front) lineup showing the same ratio during the period 1990–1993.


Figure 3
Ratio of Construction Worker Earnings to All Worker Earnings
Pre-Repeal (1976-1979) and Post-Repeal (1990-93) Periods


Source: Bureau of Labor Statistics, Employment and Earnings


• The first pairing on the left (None) is for the nine states which have never had a prevailing wage law [i.e., Georgia, Iowa, Mississippi, North Carolina, North Dakota, South Carolina, South Dakota, Vermont and Virginia].

• The middle pairing (Repealed) is for states whose prevailing wage laws were repealed since 1979 [i.e., Florida (1979), Alabama (1981), Utah (1981), Arizona (1984), Colorado (1985), Idaho (1985), New Hampshire (1985), Kansas (1987) and Louisiana (1988)].

• The other three are for groupings where the prevailing wage laws have been estimated to be of increasing severity, as measured by how closely they reflect the union wage.4 Theremaining states are grouped as follows:

Weaker (nearer-to-market rate) laws: Nebraska, Tennessee, Oklahoma [declared unconstitutional in 1995], Kentucky, Maine, Maryland, Montana and Texas;

Average laws: Delaware, Connecticut, District of Columbia, Wyoming, New Mexico, Arkansas, Indiana, Pennsylvania, Alaska, Nevada, Oregon, West Virginia and Wisconsin; and

More severe (Near-Union rate) laws: Michigan [declared unconstitutional in 1994], Illinois, Missouri, Rhode Island, Minnesota, Ohio, Washington, Hawaii, California, New Jersey, New York and Massachusetts.

None of the construction worker groups commanded wage premiums as large in the later period as in the earlier periods. This earnings decrease change occurred in states which repealed their prevailing wage laws and states which did not repeal them. (The Utah study's claim of a unique decline in repeal states is unsupported, and its failure to report the commonality of experience between repeal states and others is extraordinary in a study carrying an academic imprimatur.)

Multiple Regression Analysis. A regression model takes correlation data from various statistical series to predict the next or some future value of a related series. For example, if a correlation is suspected between height and weight in people (e.g., taller people weigh more than shorter people), one can collect data to discover if such a coefficient exists. Then at some future time, to estimate someone's weight, applying the coefficient to their height would give the estimated weight. A multiple regression model is done the same way, but with added refinements. Using a similar example, if men and women of the same height tend to weigh different amounts, with suitable data one can add sex to height and, knowing both, have a better estimate of weight. This process is repeatable, and any factor can be evaluated for relationship. The designer of the regression model determines which factors will be included. It is understood that factors not included in the model cannot impact its results.

The Utah study uses an elaborate multiple regression analysis on a data sample of 27,778 observations to predict construction wage rates in various states based on factors chosen by the authors.* These factors are:



  • region of the country;
  • number of years after an arbitrary starting point of 1975 (the "secular trend");
  • the statewide unemployment rate for private, nonagricultural workers;
  • the nationwide rate of inflation as measured by the Consumer Price Index; and
  • the status of the state's prevailing wage law—whether the state has ever had one, had one but repealed it sometime between 1979 and 1988, had one during some or all of the entire period that covered public works contracts only in excess of $500,000, or had an active one. (This last factor is called the "legal variable.")

*The number of data points, 27,778 claimed by the Utah study is puzzling, as is the collection of data on construction earnings by four-digit Standard Industrial Classification (SIC) code. Since there is no indication that inappropriate SIC codes (such as Operative Builders who build single-family housing for their own account) have been eliminated, there seems to be no reason not to use industry-wide annual averages, which would require fewer than 900 data points. The number is also puzzling because it requires that there be 32.04 SIC code classifications to reach the total number of observations (51 states x 17 years x 32.04 SIC codes = 27,778) where only 26 exist. (Office of Management and Budget, Standard Industrial Classification Manual, 1972.)

The study ignores factors such as population growth rate, annual days of sunshine, degree of urbanization, and state right-to-work status, and thus implicitly assumes that these factors do not affect construction wage rates.

The model works by starting from a hypothetical "real" annual construction wage in 1975, then controls successively for the various components. For example, Utah's 1991 average annual construction wage is "predicted" by,



  • starting at a hypothetical point of $33,005;
  • subtracting $79 because Utah is in the Mountain region;*
  • subtracting $225 seventeen times to account for the secular trend for that many years after 1975;**
  • subtracting $1,481 because there was an unemployment rate in Utah of 4.9 percent in 1991;*** and finally,
  • subtracting $1,350 because sometime during the seventeen-year period, Utah repealed its prevailing wage law.

*The range of control amounts by region is quite large, from positive $15,628 in Alaska to a negative $2,360 in the South. It ignores such items as urbanization of different states within regions.

** The secular trend amount is apparently not itself affected by regional differences, including regional inflation. Being in fixed dollar amounts, it has a much higher percentage effect on low-wage states than on high-wage ones, although no valid case could be made for why it should act this way. For 1991, the secular control variable would be .59 percent of the Hawaiian (but 1.17 percent of the New Mexican) construction worker's wage.

*** The proposed relationship between statewide unemployment rates and construction earnings is unexplained in the text.

The annual construction worker income level in Utah "predicted" for 1991 by this model was $26,266, and the income levels in all the other states could be similarly derived. But how can these "predictions" be trusted?

A bad model produces inaccurate results. Because the Utah study is a meticulous analysis using sophisticated statistical tools applied to a valid data series, some observers may feel its results should be compelling. Unfortunately they are not. There are many ways to measure the sophistication of a regression model. Two of the ways used in this article are: (a) to analyze its components and raise questions, and (b) to evaluate whether significant information was neglected. Why was some critical information not considered in the Utah study? One factor the study fails to mention is that construction workers are not paid annual incomes, but hourly wage rates, and that the two are not conformable without taking work hour changes into account. This factor is also important because of the growing evidence that nonunion employees work more hours annually than union employees.5 Another factor is the wage dynamic associated with the changing number of persons employed. Since most employees enter or leave an industry at the bottom of the wage scale, expanding workforces tend to have lower average wages, and shrinking workforces have higher average wages. This factor, though mentioned in the text, is not included in the regression equations.

But the acid test of the sophistication of a regression analysis is how well it predicts. For example, we have state-by-state unemployment rates as well as complete information on construction worker annual income per employee in 1991 dollars from the Bureau of Labor Statistics,6 so the results of the above regression equation can be compared with the actual earnings received in each state in 1991. I did this, and my results show in seventeen states the "sophisticated" multiple regression analysis gave predictions of individual average construction workers earnings that were off by amounts greater than $5,000, and the combined average variance for all states was $2,797 per employee.

This is not hopeless performance (an R2 statistic of .73 is claimed), but the question arises as to whether a more accurate prediction could be made of construction wages by changing the model used on the same data set. For example, one might note that eight of the 11 states in the South region, four of the seven states of the Mountain region, and three of the six states of the Corn Belt either have no prevailing wage law or have a repealed law. Since these are also the three regions of the country having the lowest annual wage rates in all job categories, it might be that what the model attributed to the effect of repeal of (or never having had) a prevailing wage law might simply have been further differences of regional patterns in wage rates. An alternative model that eliminated the "legal variable" (the status of prevailing wage law) and re-analyzed the coefficients assigned to regional differences might produce a better result than the Utah study's model.

A better model produces better results. To test this, I developed what I think are more accurate regional weighting factors and substituted these for the ones assigned by the Utah study, thus totally removing the "legal variable" control weights. When the results were compared, my revised model predicted 1991 annual construction wages with dramatically better accuracy. Only eight states (instead of 17) now missed their projection by more than $5,000, and the average by which all states missed their mark fell from $2,797 to less than $4. Similar results were obtained using the original and the revised models for predicting the actual wage rates of 1990, 1992 and 1993. The revised model is two to three orders of magnitude better than the original in every case examined.

The revised model gave improved results by eliminating consideration of the status of prevailing wage laws in the various states. This confirms that annual construction wages in a state have no measurable connection with the status of prevailing wage repeal independent of other forces driving regional differences in wages.*

* It is possible to have a regression model that uses only unemployment rate, secular trend, legal status, and regional location as variables, and with such a model, the least squares regression line fit to the data would no doubt produce the coefficients for legal status found in the Utah study. But these coefficients are simply the ones which give the best fit to the data for that particular model, and as we have already indicated the model used by the Utah study is of questionable validity. The Utah study also has a second regression model, using what may be a different set of another 27,778 data points (it requires different data than the first, but is said to be the same size). It is intended to show the relationship between level of construction employment per Standard Industrial Classification (SIC) code and changes in average construction wages. For unexplained reasons, this analysis uses four-digit SIC codes, assigning an average employment level of 3,540 workers per state to each code. This breakdown is strange, because the SIC code system classifies employers, not employees. For example, SIC code 1781 is for the tiny group of contractors who specialize in drilling water wells, whereas SIC code 1522 is for all the "general contractors primarily engaged in construction of residential buildings other than single-family housing" (Executive Office of the President, Office of Management and Budget, Standard Industrial Classification Manual, 1972). There is no reason to suppose that the size of these disparate groups might be affected in the same percentage or absolute way by existence or repeal of prevailing wage rate laws. Thus, to divide the industry's employees arbitrarily among these 26 [SIC code] groups of employers; and then "control" this groups' size to a national average; make assumptions about the impact of lower wage rates or the way the numbers might be increased or decreased by prevailing wage law repeal, statewide unemployment rates, and the rest; and finally expand the result back to the universe of construction workers, as the Utah study did, is simply senseless.

Utah Study Claim #2: Government Does Not Break Even

The truly shocking news from the Utah study came in its second major assertion, namely, that because Utah repealed its prevailing wage law, the state lost millions of dollars of tax revenue—more than it was able to save in decreased construction costs; $6.8 million more per year. This loss was extrapolated to the federal level where tax losses from the repeal of the Davis-Bacon Act were estimated to be $1 billion to $2 billion annually, in "real" 1991 dollars.

In determining the tax loss to the State of Utah, the authors of the Utah study used the following analytical method:



  1. They averaged construction worker incomes in states that had repealed their prevailing wage laws for a varying number of years before and after repeal, such that the sum of the years totaled 17. (There is no valid reason for performing such an average over a variable number of years, just as there is no valid reason for deducing Utah construction workers' incomes from the average wage rates of construction workers in eight other states when those figures for Utah construction workers are available directly);
  2. The income levels are all converted to "real" 1991 dollars;
  3. The differences in the averages of individual income levels in years before and after repeal (1981 in Utah's case) are multiplied by the number of construction workers in 1991;
  4. Then, to determine the total tax loss per year in 1991 dollars, Utah's income and sales tax rates are applied to this income level (after assuming all income earners are Utah tax payers and have a "marginal propensity to consume taxable items of 80 percent"—an extraordinarily high rate); and
  5. Finally, this number is converted to 1995 dollars, which, being smaller, make the losses seem larger.

It is difficult to exaggerate how bad this analysis is. Figure 4 is based on data from the Bureau of Labor Statistics and is a close approximation of a chart appearing in the Utah study. It shows the average of construction worker incomes in 1991 dollars for the nine repeal states. In the nine repeal states, the average annual incomes for construction workers declined by $1,664, from $23,348 to $21,684. The Utah study authors attribute a similar finding to repeal: "This [decrease] does not control for other factors that might have been driving down wages, but it is prima facie evidence that the repeals forced lower earnings. . . ."7


Figure 4
Average Annual Incomes of Utah Construction Workers
Pre-Repeal (1976-1979) and Post-Repeal (1990-1993) Periods


Source: Bureau of Labor Statistics, Employment and Earnings

What's astonishing is that as economists, the study's authors had to know that in the same "real" terms average annual construction worker earnings declined not only in Utah but also in almost every other state, regardless of the status of its prevailing wage law. During the same time periods, states without prevailing wage laws and states that continued to have them also showed falling "real" construction income levels (Figure 5). The average percentage decline for the states that retained prevailing wage laws is the same as the average percentage decline of those that repealed them; and the percentage decline for states that never had them is only a bit less—5 percent instead of 7 percent. The average dollar amount of the loss in states that retained their prevailing wage laws was $1,872—about $200 higher than the average loss in states that repealed them and over $700 higher than the average loss in states that never had a prevailing wage law. So if repeal drove earnings down in the nine states where repeals occurred, what would the Utah study authors have us believe drove earnings down in the 40 states where repeal did not occur?


Figure 5
Period Comparisons, Construction Earnings
In 1991 Dollars


Source: Bureau of Labor Statistics, Employment and Earnings

If we assumed that all states had the same income and sales tax rates as Utah, then the same reasoning that produced a $6.8 million tax loss in Utah because of repeal would show an average of $23.1 million of tax loss in all the states that retained their prevailing wage laws because of ... What? Non repeal?

In either case a fall in the "real-dollar" income does not result in an actual tax decrease because taxes are not paid or lost in "real" dollars. For example, assume you were in Utah in the year 1979 and, as a reward for good performance, your employer increased your salary 10 percent, from $20,000 in 1979 to $22,000 for 1980. Also assume that the combination of state income and sales taxes was 11.8 percent. How much additional tax benefit did Utah receive by virtue of your gain? The obvious answer is $236 ($2,000 x .118). The authors of the Utah study, however, would argue that your income in 1979 had been $39,000 in "real" 1991 dollars, whereas in 1980 it was only $36,300 in 1991 dollars. Therefore they would tell you that you actually lost $2,700 in "real" income despite your raise, which in turn would have "cost" the state a total tax loss of $318.60 in 1991 dollars, or $382.32 in 1995 dollars.* What nonsense?

The Utah study carried over the losses found for the repeal states and applied them equally to all 6,000,000 construction workers in the United States,** despite prevailing wage law differences. Losses are also claimed for the 107,400 new employees said to have been added as a result of repeal.*** In addition, losses are estimated for the approximately 625,000 construction workers who live in nine states without prevailing wage laws. These states were said to sustain earnings losses twice as great as in repeal states, and are projected to be impacted again by repeal, which is nonsense.

* The method of inflation used to translate 1979 and 1980 dollars into "real" 1991 dollars is the Consumer Price Index (CPI). On the basis of 1991 = 100, the CPI stood at 195 in 1979, and 165 in 1980. Following the study's example (on p. 20), we increased the purported loss 20 percent to restate it in 1995 dollars. (Actual CPI changes were 4 percent for 1991-92 and 3 percent for both 92-93 and 93-94, so the Utah study must anticipate 10 percent additional inflation for 94-95.) The study further expands the losses by imposing them on the 31,528 persons employed in construction in Utah in 1991, rather than to the smaller numbers employed in the periods either before or after repeal—27,224 and 29,723 respectively. The entire analysis for the State of Utah is repeated in slightly different form on p. 29 of the study, using the regression coefficient for repeal states of $1,477, upped from $1,350 to express it in 1994 dollars as a measure of the loss. This amount is applied as above, but to the actual number of construction employees in Utah in each year from 1987 to 1993, less a small offset. Total tax losses are balanced against a range of hypothetical savings from repeal, which depend on the value of state-financed construction. The dollar amounts of state financed construction are not adjusted to constant 1994 dollars, but are in current, and therefore smaller, dollars. Fun with numbers?

** According to Employment and Earnings, the average annual employment in construction during 1993, counting 636,518 proprietors and an indeterminate number of management personnel who would not be subject to prevailing rates, was 4,589,541. Thus no more than 3,953,023 persons could possibly be affected by prevailing rate repeal under the broadest assumptions of the Utah study. This number was readily available to the study's authors, who chose instead to use a hypothetical and unsupported number of 6,000,000.

***Yet these new employees go from unemployment or work in other industries to construction, where they are paid average annual wages of $24,000, much of which is new money on which taxes would be paid. This should certainly represent a net gain in tax receipts. Instead, the Utah study credits them with losing $1,477 each (for which government is credited with losing several hundred dollars in taxes). Even with a model as crude as this one, the new employees should be stripped from the cost side.

On the earnings-loss side of the impact calculation, the study fails to: (a) include differences in impact upon different states depending on the severity of the state law; and (b) foresee any proportionate difference in impact from repeal of the federal Davis-Bacon law in states that may or may not have a state prevailing wage law covering some or all of the same employment at rates equal to or different from those set by the Davis-Bacon Act. The impact of repeal in the study is constant from year to year, so Utah's 1981 repeal has the same constant-dollar impact on average Utah construction earnings in 2082 as in 1982.

On the government-savings side, the study is non-judgmental with respect to savings from repeal, declining to accept one estimate over another within the range of 1 to 11 percent of outlays.* In its estimates of construction costs (and therefore savings in construction costs), unlike the rest of the analysis, the study uses current rather than "real" dollars.8 Also, it grossly understated the value of relevant construction (and therefore the savings that would result from repeal) by about a factor of five.**

The Utah study does state (p. 25) that although savings may be higher or lower, the Congressional Budget Office [CBO] favors an estimate of a 1.5 percent cost savings associated with the wage effect, plus a 0.2 percent cost savings because of paperwork associated with Davis-Bacon. Where we need to illustrate a point with a concrete example, we shall use this level, but point out in passing that it is inappropriate for the Utah repeal. Unlike the Utah study's retroactive analysis of repeal impact, the CBO estimate is prospective (after 1995) because it supposes implementation of a 1983 court determination requiring rates to be set for construction helpers (a worker category found mostly in open shop construction). CBO anticipates this to occur in 1995, decreasing the excess cost of Davis-Bacon by 1.6 percentage points, and consequently lowering the savings from repeal if repeal were to take place later. (Letter of June E. O'Neill, Director, Congressional Budget Office to Honorable Cass Ballenger, Chairman, Subcommittee on Workforce Protection Committee on Economic and Educational Opportunities, April 21, 1995.) As applied to historical figures such as those used in the Utah study, the proper CBO percentage estimate would be between 3.3 percent and 3.8 percent for a law that sets rates similarly to Davis-Bacon. Prior to repeal, however, Utah set rates effectively at the union rate, so the impact of repeal, using the CBO's methodology, would be considerably higher, probably on the order of 6 percent.

** The $11.58 billion value for federal construction used by the Utah study is the figure for the amount of federally owned construction put in place in about 1985. By fiscal 1993, that figure had grown to $17.2 billion, but still massively understated the amount of construction that might be affected by repeal, because Davis-Bacon applies to state grants-in-aid as well as to direct federal spending. The total figure for 1993 is actually $47.7 billion, budgeted to rise to $55.8 billion by 1995.

Does the tax loss exceed the savings from repeal? Table 1 is a partial re-creation of a table found in the Utah study projecting the effect of a repeal of Davis-Bacon on the federal budget. It documents some of the substantial errors or fabrications employed by the Utah study to change over $1 billion of net government savings from repeal of Davis-Bacon into a loss of almost $1.5 billion—a $2.5 billion difference.


Table 1
Utah-Study Model of Actual and Reported Effect
of Davis-Bacon Repeal on the Federal Budget



  1. See text in footnote on page 21. Line numbers here correspond with those in the Utah study presentation.
  2. As discussed in the text above, there actually was no lost income except when measured in constant dollars. This is the repeal-based loss for 1991 expressed in 1994 dollars.
  3. We cannot support this number logically or statistically, but use it as given.
  4. The actual figure is the BLS-reported figure for 1993, expressed in 1994 dollars; the Utah-study figure seems to be the 1991 BLS-reported figure, expressed in 1994 dollars.
  5. Equals $1,477 x 107,400. As these new employees owe their jobs to repeal, they cannot very well be said to suffer diminished income as a result of repeal.
  6. To reiterate, there are no actual earnings decreases, therefore no decreases in tax receipts or actual tax losses. These numbers have only theoretical meaning.
  7. See the discussion in the text and previous footnote.
  8. Per Congressional Budget Office. See discussion in the text and corresponding footnote.
  9. (Line 13b – line 10.) The Utah-study figure further overstates the "loss" by not considering that, by its figures, the federal government saves less (in reduced construction costs) than it stops paying to the 10 percent of the construction workforce engaged on public works construction. The difference, $396 million in this example, represents money continued to be paid out by government for construction but not received by construction workers, so it must necessarily go to someone else. That someone else will have to pay tax. On the same assumptions as above, this tax would subtract $111 million from the government loss. It should also be noted that if the induced construction employment (said to be 107,400 here) is brought about by reducing the number of workers receiving unemployment compensation payments, and if an unemployed person receives plus-or-minus $1,000 per month in transfer payments, there is a net savings to the government of $1,655 million from the decreased unemployment. Thus, even using the erroneous Utah numbers in the table above, the net loss of $1,462 million would become a net gain to government of $304 million from repeal of Davis-Bacon.

Utah Study Claim #3: Increased Occupational Injuries

The Utah study predicts that more workers will become injured after repeal for the following reasons:



  • inexperienced construction firms will cut safety corners;
  • lower minimum wages will cause high-wage construction workers to be replaced by new workforce entrants; and
  • experienced workers will take unnecessary safety risks to keep their lower-rate jobs.

The authors claim a combined effect that is staggering:

[After repeal] there would be 76,000 additional workplace injuries in construction annually, with 30,000 of them serious and thus requiring time off from work to recover. This could lead to additional worker's compensation costs of about $3 billion per year, of which $300 million would be passed on to the federal government as increased costs on public works.9

(They also add that 130 to 150 additional fatalities per year could be expected as a result of repeal, though they do not put a price on this death toll.)10

The theoretical construct of this claim is derived from two premises: the rate of injuries decreases substantially as length of service increases; and large, experienced employers in construction have injury rates that are 80 percent lower than small-to-medium-size contractors. The first of these claims is grossly misleading, and the second is demonstrably wrong.

The implication of the length-of-service argument is that by replacing old union contractors with new nonunion ones, established employment relationships are disrupted and accidents increase. The referenced footnote authority (U.S. Department of Labor, OSHA) in the original says: "The frequency of occurrence of accidents decreases substantially as length of service increases."11 This is entirely misleading if what follows in the source is left out. It goes on:

It is unlikely that the length of service [as reported in a data element on the workers compensation form] has been interpreted as the total length of time in the construction industry. Rather, it is the length of time with the employer or on the job site where the accident occurs.12

The source implies that most accidents occur when employees are not familiar with their workmates or worksite, whereas the Utah study falsely transfers this to length of service in the industry. In fact, since union hiring practices still favor crewing up separately for each job, whereas nonunion practices increasingly favor crew continuity from job to job, replacing union workmen with nonunion ones should, if anything, result in lower injury rates. This impact, though probably small, is discernible in the statistics below.


Figure 6
Injury Incidence Rate By Employment Size
Representative Year (1978)


Source: Occupational Safety and Health Administration

Furthermore, although the Utah study's authors claim that large contractors have much lower injury rates among their employees, this claim is not supported either. Figure 6 demonstrates that injury rates actually peak in establishments of medium size, and are lower for firms at both ends of the size spectrum.13 Certainly, no competent economist could conclude from these figures that large employers have greatly reduced injury rates. Thus, both initial premises of the Utah study are seriously flawed.

Does Repeal Increase Injuries? Some aspects of the Utah study defy analysis. One example is the section comparing injury rates for construction workers—by state and by status of prevailing wage laws—based on what it says are the experiences of plumbers and pipefitters as follows:

  • the 14-year average injury rate in 32 states with a prevailing wage law is compared with that in 9 states that have never had a prevailing wage law; and
  • the same 14-year average is compared with the average injury rate in 9 states that repealed their prevailing wage laws between 1979 and 1988: (a) for 1 to 10 years before repeal, and (b) for the 3 to 12 years after repeal.

The difference noted between "before repeal" and "after repeal" was said to be the amount by which injury rates would increase in the whole United States after the repeal of Davis-Bacon, which, when applied to the estimated cost (in an unspecified year) for workers' compensation in the construction field yielded a $3 billion estimate for the cost of repeal.

Although this is complex analysis, it is not sophisticated analysis. First, it needlessly bases its conclusions on a subset of the construction workforce. The Utah study says that plumbers and pipefitters (Standard Industrial Classification [SIC] 171) were chosen because this specialty trade has injury rates in the mid-range for construction. This is not supported by the U.S. government sources cited, where this category is typically the second highest of all 3-digit codes within the construction industry. Furthermore, since all construction workers are presumably affected by prevailing wage laws, it is more reasonable to evaluate the injury and illness rates of all construction workers than to deliberately measure a non-typical subset's patterns and apply them to all. Second, it mentions but does not control for differences in injury rates depending on the type of work done.* Third, it draws inferences from direct comparisons of data sets taken from variable time frames. Fourth, it generalizes its statistical result inadequately. Fifth, it improperly correlates injury rate results with workers' compensation spending estimates to determine impact.** These are not the only problems.

* If one were to assume there actually was an impact on injury rates as a result of inexperienced or untrained workers taking over prevailing rate work after repeal, there would be no reason to suppose a proportionate impact on segments such as road building, where plumbers and pipefitters are nominally represented, as on commercial construction, where they are numerous. This is a different statistical problem from that of extending the injury rate results from specialty contractors (plumbers and pipefitters) to other construction segments to other industrial segments, like operative builders or residential contractors, whose employees or employees activities would be very little influenced by prevailing rate repeal. Both analyses are neglected by the Utah authors.

** The Utah study authors talk about injury rates, serious injury rates, and days lost per year but show no apparent appreciation of the difference between these series and workers compensation statistics. Accident rates reported for workers compensation purposes (by only ten states which provided this data to the Bureau of Labor Statistics each year for the four consecutive years of the study) are quite different from accident rates reported to OSHA, partly because states delay from one to eight days before classifying an injury or illness as serious and reporting it. (U.S. Department of Labor, Occupational Safety and Health Administration, Construction Accidents: The Workers Compensation Data Base 1985-1988, April, 1992.)

The data the Utah study claimed to use are not actually available in the source materials cited there.* In fact, according to a spokesman for the Occupational Safety and Health Administration, the only published data on construction industry injury rates by state for this time period are those depicted here in Figure 7—but even this data is based on incomplete information because not all states participated in the survey.** The graph below shows the incidence of injuries per 100 workers in the construction industry for the years 1975 to 1978 for three groups of states:

* The Utah study's source, as cited in its text, p. 62, is the BLS Occupational Injuries and Illness Survey report for various years. But these reports do not give state-by-state data breakdowns, and have never done so. A special Occupational Injuries and Illness by State report was issued in 1980, giving data for 1975-76, but there has been no published report since. Furthermore, because of varying state participation in the survey, those state-by-state data that were available did not contain reports for several key states, including New York, New Jersey, Illinois, Ohio, Texas and others. Spokesmen at the OSHA Office of Statistical Studies and Analysis, interviewed July 29, 1995, stated that full state-by-state information is not available and has never been available in any form, either for plumbers and pipefitters or for construction workers generally.

** From the OSHA hearings, 1980. Georgia, one of the never-had states, declined to participate in the survey, as did Illinois, Ohio, New Jersey, New York, and Texas, all of which did have laws. Additionally, data are missing for Nevada (1975), Michigan (1976), Arkansas, North Dakota, and New Hampshire (1977 and 1978) and Oklahoma, Colorado, North Dakota, Tennessee, and Mississippi (1978). It does not appear that data missing from individual years should affect the results.

  • those that never had a prevailing wage law;
  • those that would repeal their prevailing wage law during the following decade; and
  • those that did have prevailing wage laws.

These figures yield exactly opposite conclusions to those reached by the Utah study for the pre-repeal situation.


Figure 7
Injuries Per 100 Workers, Construction Industry
By Status of Prevailing Wage Law, 1975-1978


Source: OSHA Oversight Hearings, 1980

Regarding construction-related injury rates, the authors of the Utah study said,

Injury rates in construction were relatively low in the nine repeal states prior to repeal. . . . In the 32 states that have retained prevailing wage laws, injury rates have been and remain relatively low. In nine states that have never had state prevailing wage laws, injury rates were and remain relatively high.* 14

* For reasons already stated, it is highly unlikely that the Utah study could possibly have reported data for 32 states that retained prevailing wage laws, since data was available from the stated source for a maximum of 27 states. Also note that the chart (Figure 4.1, p. 63) is titled "Injury Rates In Construction," whereas the text speaks of injury rates to plumbers and pipefitters, SIC 171. We can neither confirm nor deny the injury rates presented if they are for plumbers and pipefitters, since we know of no possible source for that information. It seems unlikely, however, that the pattern for the injury rate for this subset would be the inverse of the pattern for construction workers generally.

In contrast to their assertions, Figure 7 clearly shows the following about average injury rates:

  • they were the highest of all for pre-repeal states;
  • they were, in the pre-repeal time period, relatively low for states that never had a prevailing wage law; and
  • they were relatively high for those that did have a prevailing wage law.

Figure 8, drawn from unpublished OSHA data, plots the average incidence of injury in representative time periods before and after repeal.* Obviously, in repeal states, injury rates did not increase after repeal. States with prevailing wage laws (which tend to be states with more union employment) continue to have higher injury rates than states that never had a prevailing wage law. Also, the states that repealed their prevailing wage laws still have the highest injury rates.

*The pre-repeal data, for 1975 through 1978, are from the "Oversight Hearings on OSHA Occupational Safety and Health for Federal Employees, Part 4: State Plans." 96th Congress, 2d Session, March 4, 11, 19; April 1, 29; May 28; June 17, 25; July 22; and September 16, 1980, p688. This gives incidence rates fro construction injuries and illness per 100 full-time workers in 45 reporting states in the five years, 1974-78. It is the greatest concentration of published state-by-state data on construction injuries. In addition, data for 1990 through 1992 were made available to the author (in Washington, D.C., July 29, 1995) by Joseph DuBois, Director of Data Analysis, U.S. Department of Labor, Occupational Safety and Health Administration. Mr. DuBois, who had also prepared the data used in the 1980 OSHA hearings, verified that both data sets were similarly derived from individual state submissions, using the same definitions and statistical methodologies. He also confirmed that data are not available in any format from the federal government covering construction worker injury rates in Georgia, Illinois, New Jersey, New York and Ohio for either time period; or from Texas in the earlier period; and South Dakota, Colorado, Idaho, New Hampshire, the District of Columbia, Pennsylvania and Wisconsin in the later one. For comparison purposes here, if no data were available for either time period, no estimates were made. If data were available for one or the other period, I assumed no change in average rate between periods.

Figure 8
Change In Average Construction Injury Incidence
Pre-Repeal and Post-Repeal Periods
By Status of Prevailing Wage Law

Source: Unpublished files, U.S. Department of Labor, Occupational Safety and Health Administration

Why did the Utah study make incorrect conclusions about construction-related injury rates? Perhaps because the authors used another faulty multiple regression analysis model. This model, which does not predict injury rates but the log of injury rates, has four components: (a) a time trend, (b) a regional variable, (c) status of a prevailing wage law, and (d) the unemployment rate. It has a low to very low R2 statistic (.16 to .49), and the following is an example of how it works:

  • If in the years before prevailing law repeal, states within the same region (no regional control variable),
  • and the same time frame (no difference in secular trend),
  • had the same unemployment rate (which, within the regions is an acceptably close estimate),
  • then their injury rates (or the log of their injury rates) will differ from one another only based on status of prevailing wage law.

Thus, they asserted, states which never had a prevailing wage law should show higher injury rates than states with a prevailing wage law (even if they later repealed it). The actual statistics show, however, that such is not the case.

The discrepancy between the Utah study's model and actual statistics can be seen by examining the data from just one region of states. The Bureau of Labor Statistics' Region IV comprises four states which never had laws, and four which had them in the period 1975–1978. The Utah regression model says the "never had" states should have higher injury rates. Here are the actual injury rates:


Table 2
Construction Injuries
BLS Region IV, 1975-1978

Source: Unpublished OSHA data in author's posession

Clearly, the injury rates in "never had" states are lower.* In all cases where statistics are available to verify this model, the model works backwards.

*Similar results are obtained in Region VIII, where North and South Dakota are substantially below their "have law" counterparts in Wyoming, Colorado, Montana and Utah.

Since there is no increase in injury rates in repeal states, there is no cost associated with repeal. Neither are there additional deaths. Based on the statistics above and on union and nonunion death rates in construction (the union fatality rate is consistently three to six percentage points higher than the nonunion rate),15one could assign injury-related cost differentials to prevailing wage states, thereby estimating a cost to the country of maintaining these laws, but this would be as meaningless as the Utah study's conclusions (although better supported by the facts).

The conventional wisdom, that injury and death rates in construction are independent of union membership and of prevailing wage status, is reaffirmed. Specific documentation of this finding can be obtained in OSHA's Analysis of Construction Fatalities – The OSHA Data Base 1985–1989.*16 which gives the following statistics:

* OSHA's Analysis of Construction Fatalities notes that the distribution of fatalities among union and nonunion worksites is similar to the composition of the construction workforce in terms of union and nonunion workers for the 5-year period. It also notes (p. 19) that "The percentage of fatalities for the various size construction firms is similar to the representation of the construction workforce in terms of firm size."

Table 3
Fatalities on Worksites, 1985-1989

Source: OSHA Analysis of Construction Fatalities (cited above)

Thus, nonunion workers have a smaller rate of fatal injuries than their percentage of the workforce; and union workers have a higher rate of fatal injuries than their percentage of the workforce. Although there were 2.3 times as many fatalities on nonunion sites as on union ones, this is less than what would be expected, since there are 3.5 times as many nonunion workers. Thus, all the conventional wisdom about repeal of prevailing wage laws of the states or repeal of the Davis-Bacon Act is reaffirmed.

IV. New Estimates of Savings Likely From Repeal of Davis-Bacon

The primary purpose of this article has been to examine the irrational conclusions of the Utah study, which claimed that repealing the prevailing wage laws:

  • drives construction wages down below market;
  • costs governments more in lost taxes than they save in reduced construction costs; and
  • increases workplace injuries, deaths, and related expenses.

In contrast, I found that the facts support these conclusions:

  • repeal lowers artificially inflated wage rates but only towards the market level, not below it;
  • lower construction worker earnings and government tax losses after repeal exist only when measured in theoretical constant dollars; when measured in actual current dollars, they are really increases;
  • savings from prevailing wage law repeal are both real and substantial; and
  • construction injuries are actually higher in states that retained their prevailing wage laws than in those that do not have such laws.

In the mid-1980s, I concluded that although the amount of savings from Davis-Bacon repeal is unknown because of the different ways in which prevailing wages are calculated, the majority of reliable estimates clustered around $1 billion a year—savings to the federal government of from 1 to 3 percent on some projects, up to 4 to 7 percent on others.17

Various approaches were followed to reach these estimates, which I won't repeat here, but a common problem among approaches based on labor cost differentials is establishing the level of wage rates actually supported by the prevailing rate process so we could tell how much they might drop on repeal. Gauging wage differentials between union rates and market rates is possible (e.g., in 1983 the differential was 1.72, and in 1993 it was 1.62),* but the Davis-Bacon Act, unlike certain state prevailing wage laws, sets union rates only part of the time. Sometimes it sets minimum wage rates or survey-average rates. Thus, estimates of savings from repeal are clouded by uncertainties about how closely Davis-Bacon rates approached union rates to start with. In evaluating the Utah study, I found a new statistical avenue around these problems.

*Professor Philips supposes (Wages and Benefits . . . undated, fn. 4): "The historic union-nonunion wage rate differential in the United States has remained fairly constant at around 20 percent." The Congressional Budget Office, to the contrary, states: "The ratio of cash wages for union construction workers to those for nonunion construction workers was 1.62 in 1993, as compared with 1.72 in 1983." (Letter from June O Neil, Director, Congressional Budget Office, to Honorable Cass Ballenger, U.S. House of Representatives, April 21, 1995, p. 3.) Adding in fringe benefits would make the ratios even higher.

The analysis begins with the hypothesis that repeal of a state's prevailing wage law moves average earnings of construction workers toward earnings of all the state's other workers.* Wide variations exist in the average earnings of workers in different states. Some of these differences have to do with climate, historical patterns, regional growth, population density, urbanization or union strength. Assume that whatever they are, the wage differences would affect workers in the construction industry to about the same degree that they affect workers in other industries. Then if there are higher relative construction wages in states with stronger prevailing wage laws, the savings from eliminating those laws and allowing local market conditions to set wage rates (just as it does in other industries) could be estimated.

* The wide variations in earnings from state to state are preserved in this model. In 1993, the lowest average weekly wage was reported by the Bureau of Labor Statistics (in Employment and Earnings, 1993) for employees in South Dakota, at $358 per week. The highest, excluding the special case of the District of Columbia ($754 per week) was Connecticut, 1.8 times higher at $638 per week. For construction workers, the lowest was found in Arkansas ($395) and the highest in Alaska ($846).

Table 4 shows this first step. It is a restatement of the data already presented in visual form in Figure 3. During the "Pre-Repeal Period," the ratio of construction worker earnings to all worker earnings increases corresponding to the strength of the state's prevailing wage law—from a low of 1.18 in states that never had a prevailing wage law, to successively higher ratios for states whose prevailing wage law is characterized as weak, average, or strong (likely to produce near-union rates). In the "Post-Repeal Period," the ratio found in the repeal states closely approximates that in the states that never had a prevailing wage law—a finding consistent with my hypothesis.


Table 4
Ratio of Construction Worker Earnings
to All Worker Earnings

Source: Bureau of Labor Statistics, Employment and Earnings,various years.

To take the analysis to the national level, let us assume that the impact on a state's construction earning levels from repeal of the Davis-Bacon Act would be the same as the impact from repeal of the state's prevailing wage law.* Then the impact of repeal on government spending will be in proportion to the state's construction work done for different levels of government. So, if federal and joint federal-state projects are 20 percent of a state's total construction work, then 20 percent of the savings from repeal would accrue as reduction in federal costs.

*Although there is no way to verify this short of actual repeal, there is no reason to believe it would not be true. Indeed, most state prevailing wage laws would probably follow a repealed federal law into oblivion. Furthermore, the observed relative earnings impact in repeal states occurred as though all public works jobs were affected, although federal projects and joint federal-state ones continued to be constructed under Davis-Bacon rates, so the opposite is also likely to occur.

Figure 9 is determined from Table 4 as follows: For each state within the category of weak prevailing wage laws, the additional ratio reduction (1.10 – 1.08 = .02) is multiplied by,

  • the average number of construction workers in those states for the same time period (1991–93, not shown), and by,
  • the average annual earnings of the state's construction workers.

Savings for individual states are summed for the sector total. The process is repeated for the other two sectors, and the grand total is presented.

Figure 9
Sector and Total Savings
From Repeal of Prevailing Wage Laws

Source: See text.

Federal and federally aided construction subject to Davis-Bacon amount to 21 percent of the national construction market,* so the federal government might expect to save $1.3 billion of the 1993 budget, or $1.58 billion on the larger, 1995 federal budget. If administrative savings were also included, federal savings would top $1.8 billion a year—substantially higher than previously estimated but still in the 3 percent range of total project cost. However, if the state and local governments follow the lead when the federal Davis-Bacon Act is repealed, the total savings at all levels of government may exceed $4 billion per year, counting administrative and enforcement spending, since public works account for about 56 percent of all construction spending.

*According the Bureau of Economic Analysis, U.S. Department of Commerce, the national income without capital consumption adjustment of the construction industry in 1993 was $228 billion. The International Trade Administration of the U.S. Department of Commerce reports federal and federal grant outlays for construction in fiscal 1993 to be $47.2 billion. Hence, Davis-Bacon would cover 21 percent of the total. This is not evenly distributed state by state, but should be roughly proportional to the number of construction workers in each state.

V. Conclusion

The Utah study set out to overturn the conventional wisdom about the positive economic impact that would accrue from repeal of the Davis-Bacon Act and state prevailing wage laws. It is an impressive-looking piece of work, but fundamentally wrong in many ways. All of its conclusions that I evaluated are wrong, and the conventional wisdom that it sought to refute is not only supported, it is reinforced by the new statistical approach developed here. More than ever before, I believe that the Davis-Bacon Act (whose cost to the federal government alone may exceed $1.8 billion per year in excess government spending) should be repealed.

*Dr. A. J. Thieblot is a consulting economist residing in Baltimore, Maryland.Back to Text




1 For a full explanation of the Davis-Bacon Act and other prevailing wage laws, see inter alia, A. Thieblot, Prevailing Wage Legislation: The Davis-Bacon Act, State Little Davis-Bacon Acts, The Walsh-Healy Act, and the Service Contract Act, Labor Relations and Public Policy Series No. 27 (Philadelphia: Wharton Industrial Research Unit, University of Pennsylvania, 1985), p. 1-43; and A. Thieblot, State Prevailing Wage Laws, An Assessment at the Start of 1995 (Rosslyn, Va.: State Relations Department, Associated Builders and Contractors, 1995), p. 1-39. 

2 P. Philips, G. Mangum, N. Waitzman, and A. Yeagle, Losing Ground: Lessons from the Repeal of Nine "Little Davis-Bacon" Acts (University of Utah, February 1995). Referred to as "the Utah study."

3 These figures were provided by the State of Utah, Department of Community and Economic Development, citing in turn the Utah Department of Employment Security, Labor Market Information and Research. For the period between 1953 and 1993, they cover average monthly employment, average monthly wages, and average number of establishments in the first quarter for all employees on nonagricultural payrolls and for various combinations of 2-digit Standard Industrial Classification (SIC) codes in the state. (Data in the author's possession.)

4 Explanations for the listings of increasing severity are found in Thieblot, State Prevailing Wage Laws (1995), pp. 12-18. 

5 See H. Northrup, Open Shop Construction Revisited(Philadelphia: Industrial Research Unit, The Wharton School, University of Pennsylvania, 1984), p. 520. 

6 Unemployment rates are found in U.S. Department of Labor (DOL), Bureau of Labor Statistics (BLS), Geographic Profile of Employment and Unemployment, 1991. (Also available for other years.) Earnings information is from DOL, BLS, Employment and Earnings, 1991. This series is available, sometimes called Employment and Wages, for a number of years, some exclusively on microfiche. As of this writing, a few years from the mid-1980s do not seem to be available in any distributed format. Their lack is not critical to the analysis.

7 The Utah study, using data from 1975-1991, found the nine repeal states average incomes to decline from $24,317 to $22,482, a difference of $1,835 (Figure 2.5, p. 19.). We do not consider the variation from the numbers in the text significant.

8 See Utah study, p. 30. Also, see U.S. Department of Commerce, International Trade Administration, Construction Review, Spring, 1994, Table A-3. 

9 From the Utah study executive summary. Curiously, in testimony prepared for delivery to the Ohio Legislature in June 1995, Professor Philips (the lead author of the Utah study) said, "We found that injury rates rose by 15 percent after the nine states repealed their prevailing wage laws. We did not calculate the increased worker compensation costs these injuries created nor the cost to the state for caring for uninsured construction workers." P. Philips, Wages and Benefits as a Percent of the Net Total Costs in the Construction Industry (Salt Lake City, Utah: Economics Department, University of Utah, undated), p. 6.

10 Utah study, endnote 82.

11 U. S. Department of Labor, Occupational Safety and Health Administration, Construction Accidents: The Workers' Compensation Data Base, 1985-1988, April 1992, p. 13, cited in Utah study at endnote 78.

12 Same source page.

13 U.S. Department of Labor, Bureau of Labor Statistics, Occupational Injuries and Illnesses in the United States by Industry, 1978, Bulletin 2078, August 1980, Table 6. Although the numbers shown in Figure 6 are for 1978, there should be no material change from year to year.

14 Utah study, p. 63.

15 C. Culver, Construction Fatalities: Comparison of Nonunion and Union Contractors (Rosslyn, Va.: National Center for Construction Education and Research, 1995). Table 4 gives comparison of union and nonunion contractor fatalities for the years 1985-1993. 

16 For death rate statistics, see OSHA's Analysis of Construction Fatalities - The OSHA Data Base 1985-1989, p. 18, notes. 

17 See Thieblot, Prevailing Wage Legislation (1985), p. 93-113