The Biden administration recently proposed substantial revisions to the Davis-Bacon Act. Originally passed during the New Deal, the Davis-Bacon Act mandates that contractors receiving federal funding pay their workers at “prevailing wage rates.” This new proposal aims to redefine the ‘prevailing wage,’ reversing adjustments implemented during the Reagan administration.
Under the Davis-Bacon Act’s current framework, the Department of Labor conducts a series of surveys of federal contractors to estimate a given industry and location’s prevailing wage. If the survey shows that more than 50% of workers are paid a particular wage, that becomes the “prevailing wage” for subsequent government contracts in that industry. If not, the prevailing wage is simply the average of all reported wages.
The Biden administration proposes changing this two-step process to a three-step one, incorporating an additional intermediate step. If there is no wage paid to at least 50% of workers, the next step would be to check to see if a given wage rate is paid to at least 30% of workers, which would become the “prevailing wage.” As before, if no single wage is paid to 30% of workers, the average wage would be used. The change would restore Davis-Bacon procedures to those that were used from 1931 to 1982.
Most of the discussion about this rule has focused on the potential of the rule to raise contractor wages. However, this is somewhat overstated – even when the new 30% rule is applied, it would rarely be substantially higher than the average wage. More importantly, the specific algorithm used is almost irrelevant because of the weaknesses of the survey data in the first place.
One problem is that most contractors do not complete the survey, either because they don’t think it’s worth their time, or because they don’t even see the survey in the first place. A 2011 Government Accountability Office report found that almost half of prevailing wage rates were based on a survey of a dozen or fewer workers. Moreover, the few contractors that do fill out the survey could be those that would benefit the most from higher rates, biasing the survey results upwards.
Another problem is that the data is often years out of date. A 2019 audit of Davis-Bacon surveys showed that it typically took two and a half years to complete a survey. As a result, published prevailing wages substantially lagged behind the actual labor market. In some cases, the contract rates were decades old. With median construction wages increasing by roughly 4% annually, “prevailing wages” can easily be 10% or more behind the actual labor market.
Both of these issues mean that the prevailing wage assigned for different industries will not reflect the actual wages being paid to contractors. Rather than meticulously aligning government rates with local industry standards, we’re rolling the dice and hoping for the best. In some cases, this may result in government wages that substantially exceed the actual prevailing wage. In other cases, it may be substantially less.
The Biden administration’s proposed changes do include measures to address these issues – for example, by allowing prevailing wages to be determined by surveys conducted by state or local governments in cases where there is no recent data from the federal surveys. Another provision allows the use of Employment Cost Index (ECI) data from the Bureau of Labor Statistics to update prevailing wage estimates that are over three years old. The ECI uses a larger and more statistically rigorous survey to estimate labor costs, which is a substantial improvement. Still, three years is a long time in the labor market.
In a world where Davis-Bacon data was based on accurate and timely surveys, a debate about the precise formula used to determine prevailing wages could be well worth having. But that is not the world we live in. Instead, we live in a world where much of our data is outdated.
Rather than increasingly complicated algorithms, we should look for more straightforward measures to determine contract rates. Ideally, Davis-Bacon laws could be eliminated altogether because of stronger labor laws that help keep contractor wages high. This also has the appeal of not uniquely applying to contractors who work for the government, which makes building vitally important public infrastructure more expensive than private sector work.
In the absence of such laws, why not simply use the ECI to determine prevailing wages in the first place, instead of merely as a backstop for Davis-Bacon when it fails? We have to ask why the government should do two expensive surveys – especially when the flaws of the Davis-Bacon surveys are so pervasive.
It is crucial to establish government rate policies that acknowledge the flaws in our data rather than disregarding their existence. We need to do so before government contract rates will continue to be determined by an arbitrary and unfair process.