Ed Hudgins, director of regulatory studies at the conservative Cato Institute, added that the survey also failed to take non-tax benefits, such as vacation time and retirement savings plans, into account. And he added that the survey relied on self-reporting, a methodology technique that leads to a relatively high number of errors.
Martha Farnsworth Riche, an economist and former director of the Census Bureau who vetted the study for the accounting office, acknowledged that the data used by the study do not take years of experience into account, which employers often use to determine salary levels. She called for further analysis of more detailed data from the 2000 Census to assess how education and years of experience affect salary differentials.
Farnsworth Riche added, however, that the broad definition used to define management positions might have skewed the results in the opposite direction.
"By looking at the industries that employed nearly three out of four women managers, and by having a broad definition of management, we definitely did not come up with as dramatic results as we could have," she said.