Apples to Oranges: Comparing Prevailing Wage Rates
Earlier, I showed how it's nearly impossible for the prevailing wage to add 25% to the cost of public construction projects, like opponents to the law claim, even if you assume that the prevailing wage is nearly 50% higher than average wages in the construction sector. But let's take a closer look at that 50% claim.
Opponents to West Virginia's prevailing wage law claim that the way the prevailing wage is calculated is flawed, and biased towards higher wages. This claim is supported by a study from the conservative Public Policy Foundation of West Virginia, which found that West Virginia's calculation for the law results in prevailing wages that are 49% higher than the average market wage. However, the Public Policy Foundation's method of comparing wages was itself flawed.
The study compared the prevailing wage rates set by the West Virginia Division of Labor with the average wages for construction occupations from the Bureau of Labor Statistics Occupational Employment Statistics (OES). And prevailing wages are an average of 49% higher than the averages for construction occupations from the OES.
But while using the OES data allows for comparisons of occupations, the OES is measuring a different portion of the construction industry than what receives the prevailing wage. The OES data includes workers in the residential construction sector, which typically employs workers with lower skill levels and less experience than those who work on large projects typically funded by the state. Prisons, schools, and bridges are larger, more complex projects than what is found in the residential construction sector, and require more skilled and experienced workers. As a result, the workers are higher paid.
In fact, when you compare wages for residential and nonresidential construction workers, which you can do using data from the Quarterly Census of Employment Wages (QCEW), nonresidential workers do make far more. The gap between the average wages from the OES and the prevailing wage rates set by the Division of Labor reflects the gap between residential and nonresidential construction.
And it's not just West Virginia's prevailing wage law that is creating the wage gap between residential and nonresidential construction. Virginia, with no prevailing wage law, has a similar wage gap. And while Virginia's gap is less pronounced than West Virginia's, that is due to its residential wages being higher than West Virginia's, rather than West Virginia's nonresidential wages being excessively inflated by the prevailing wage.
Using OES data to compare prevailing wage rates overstates the cost of prevailing wage. While the point of the prevailing wage law is to ensure that wages paid to workers on state construction projects don't result in a race to the bottom, they probably wouldn't fall by 50% without the prevailing wage. And even if they did, the state wouldn't save 25% on the cost of public projects like prevailing wage opponents claim.
Does that mean that the wages from the QCEW should be used to set prevailing wage rates? Not really. The wages in the QCEW are the average of all workers at a particular establishment, which can include multiple occupations and jobs. So while the QCEW shows that workers at a roofing establishment earn an average weekly wage of $928 in West Virginia, those workers aren't all necessarily roofers. That wage figure includes managers, receptionists, and other non-roofing occupations that can exist in a roofing establishment.
In addition, the average weekly wage in the QCEW is calculated by dividing the annual wage by 52 (weeks). The construction industry is very seasonal, with most workers working less than 52 weeks in a year (according to the Current Population Survey, construction workers in West Virginia work an average of 47 weeks per year). This makes their average weekly wages lower in the QCEW, since it includes several weeks of not working. So a construction worker earning $25/hour, but only working 47 weeks/year would show up in the QCEW as earning $904/week or the equivalent of $22.60/hour. And since the prevailing wage is set hourly, you really can't compare the two.
And if the OES data isn't an appropriate comparison, and the QCEW isn't an appropriate comparison, what can you do? You can take a survey, which is exactly how West Virginia's prevailing wage rates are set.
If West Virginia's prevailing wage rates really were that far out of line with the market averages, then our construction costs should be higher than in other states, particularly those without a prevailing wage law. But, that is not the case. This study, prepared for a county government in Maryland, looked at the median cost of school construction per square foot for six mid-Atlantic states, both with and without prevailing wage laws. And it found that West Virginia had the 2nd lowest construction costs of the six states. Our per square foot construction costs were only 0.7% higher than North Carolina's, the state with the lowest costs. Our construction costs were 20% lower than in neighboring Maryland, which has only partial prevailing wage coverage, and 7% lower than neighboring Virginia, which has been cited as a model by opponents of West Virginia's law.
While the prevailing wage does result in higher wages, those higher wages result in more experienced workers with greater productivity. And that productivity offsets the higher wage. That's why, as the study found, while the prevailing wage created a wage premium for construction workers, "there is no measurable or statistically significant increase in construction costs associated with prevailing wage regulations." The report also found that it did appear that prevailing wage laws encouraged the use of local contractors.
So, once again, opponents of West Virginia's prevailing wage law have overstated their case, and their math doesn't add up. In reality, the impact of West Virginia's prevailing wage law is consistent with its original intent: to promote a path of high-wage economic development by keeping our tax dollars in West Virginia.