Michael J. Hicks, PhD, is the director of the Center for Business and Economic Research and the George and Frances Ball distinguished professor of economics in the Miller College of Business at Ball State University. His column appears in Indiana newspapers.

In 1931, the federal Davis-Bacon Act mandated that wages paid to contracted workers on government construction projects be set at the prevailing level. That ‘prevailing wage’ effectively ended competition on salaries for businesses bidding for federal government construction jobs.

The most likely reason for the initial ‘prevailing wage’ laws is that the legislation favored local unions and prevented low-cost bids from newer businesses. The great migration of African-Americans to the North was well underway in 1931, and there was widespread fear that this increased supply of workers would suppress wages.

Over the years, several states passed their own prevailing wage laws, setting a high bar for wages on construction project costs. Not surprisingly, many citizens were unhappy to pay an elevated price for government infrastructure, such as schools, roads and bridges.

Over the last three decades, a number of studies have measured different effects of these prevailing wage laws. Most found that these artificial wage and salary rules caused the cost of infrastructure to rise. A few found no effect, and many other studies looked at non-wage effects, such as the amount of job automation, the racial mix of employees or workplace safety.

The challenge of all these studies is that there were very few changes to state prevailing wages laws. So, without the change needed to take advantage of a natural experiment, none of these studies could really provide results that convinced enough policymakers to act. So, prevailing wages continued on like any other classic government giveaway program. The costs are borne by everyone, while the benefits were concentrated just to a few. There are a few loud, well-financed supporters who benefitted, while the rest of us largely suffered in silence.

Fortunately, between 2015 and 2018, six states eliminated prevailing wage legislation. Indiana was the first of these, followed by West Virginia, Arkansas, Kentucky, Wisconsin and finally Michigan. The timing of these changes offered sufficient variation for an economist to construct a natural experiment to test the effect of repealing the state prevailing legislation. My study of the repeal of prevailing wage was published earlier this month.

My work focused on two questions. First, did repealing state prevailing wages reduce the cost of road construction and maintenance projects? Second, did repealing state prevailing wages shift the mix of workers and equipment used on these projects?

To conduct this study, I collected data on road costs and road conditions from federal sources. I wanted to control for the effects of annual changes in road spending, recessions and border effects for jointly funded projects. There are wage trends and regional variation in wages for construction workers. The number of vehicle miles travelled also affects road quality, which I used to modify my measure of costs. It takes fewer tax dollars in the short-run if you allow roads to deteriorate.

All of these varied by year. I also used variables that controlled for factors I may have left out, like weather or climate that varies state by state or year by year. There were a few more challenges. In the wake of Hurricane Katrina, federal prevailing wages were suspended briefly, and in the years thereafter, Louisiana saw an unusually large amount of road work.

State laws also varied, with different cost thresholds for prevailing wages. Thus, the lack of uniformity made measuring a law as merely a ‘0’ or ‘1’ value in a statistical model a concern. Also, some states have a lot higher share of federal roads than others, and, except for a few weeks in 2005, the federal prevailing wage law was in force. The problem with this is that if the federal share is high, the effects can spillover into non-federal projects.

I ran several models, just to ensure that adding or removing one of these variables didn’t overly influence the results. Across the board, my models showed that repealing state prevailing wage laws saved taxpayers money. The savings ranged from 8.5 percent to 14.3 percent of road construction and maintenance costs.

The saved costs per mile were notably large. Here in Indiana, ending the state prevailing wage law in 2015 saved Hoosier taxpayers between $4,424 per mile and $6,866 per mile. Arkansas saw the lowest savings, with a lower threshold of $3,122 per mile, while Wisconsinites saved as much as $15,682 per road mile. My findings were pretty much ‘middle of the road’ compared to earlier studies that found cost reducing benefits of ending prevailing wages.

The second part of my study examined whether prevailing wage shifted the mix of workers and equipment on road projects. Economists measure this by calculating the labor share of construction spending. The presence of prevailing wage would tend to reduce the demand for workers and cause businesses to substitute more expensive equipment instead of paying a wage premium.

Using similar techniques, I was unable to conclude that the prevailing wage law reduced the labor share of road construction spending because the effect, though large, was just outside of the commonly understood levels of statistical significance. This is a topic worth revisiting in a few years because this measure would adjust more slowly than overall costs.

In the months since I started this study, Michigan has again changed their law, and now has a state prevailing wage. I am eager to test the effects of this change, and I’m pleased that citizens of another state will provide tax dollars for this natural experiment.

I already noted the political economy of prevailing wage is to concentrate benefits but diffuse costs. This ensures an animated, vocal group of supporters but less opposition to prevailing wage. But, there’s another more important argument about its merits, or lack thereof.

Suppose that we thought there were a group of workers who needed more government support than free markets gave them. So, we wished to transfer some tax dollars to them to improve their earning capacity, or maybe we wanted to subsidize their continued schooling. These are policies that most citizens have long supported.

The prevailing wage law actually reduces available tax dollars for these programs. A state prevailing wage law diverts tax dollars away from critical infrastructure, and away from programs that would help the most needy. It sends tax dollars away from places they are needed to places where they are not. So, if you believe your state government should spend more on people, the prevailing wage law is among the worst ways to do so. Hoosiers are fortunate indeed that the legislature ended its wasteful prevailing wage legislation back in 2015.