Teachers are actually overpaid by at least 50 percent of their fair-market value, the paper argues.

A new policy paper from the Heritage Foundation, a conservative think tank that espouses free-market principles, argues that—contrary to popular opinion—teachers aren’t underpaid. Instead, the paper says, teachers are actually overpaid by at least 50 percent of their fair-market value, costing American taxpayers more than $120 billion each year in excessive labor costs.

Produced in partnership with the American Enterprise Institute for Public Policy Research (AEI), another research institution that champions conservative ideals, the paper compares the combined average wages and benefits of public school teachers with those of private-sector professionals who have similar skills. It concludes that wages for the two groups are about the same, while benefits and job security are significantly higher for teachers.

“Teacher compensa­tion could therefore be reduced with only minor effects on recruitment and retention,” wrote the paper’s authors, Jason Richwine and Andrew G. Biggs. “Alternatively, teachers who are more effective at raising student achievement might be hired at comparable cost.”

Richwine is a senior policy analyst at the Heritage Foundation’s Center for Data Analysis, and Biggs is a resident scholar at AEI.

Using data from the Census Bureau’s Current Population Survey, Richwine and Biggs found results that are consistent with other research: Public school teachers earn about 19 percent less in wages, on aver­age, than non-teachers with the same level of education.

But then, in a surprising twist, the researchers argue that advanced degrees for teachers aren’t as rigorous as those earned by private-sector employees—and so a better comparison would be to look at cognitive ability rather than years of education. (Editor’s note: See “Four fallacies of the ‘teachers are overpaid’ argument.”)

To do so, they examined scores on the Armed Forces Qualification Test (AFQT), a cognitive exam similar to an IQ test. They reportedly chose this test because the National Longitudinal Survey of Youth linked AFQT scores to other individual-level data, so they could do a regressive statistical analysis of AFQT scores along with other control variables. Participants in the longitudinal survey were paid to take the AFQT, so the scores weren’t restricted only to people who served in the military.

Richwine and Biggs claim that “the wage gap between teachers and non-teachers disappears” when public school teachers are matched with private-sector employees who have similar AFQT scores.