The Government Accountability Office (GAO) recently released yet another report on the costs associated with the “Don’t Ask, Don’t Tell” (DADT) law, which effectively required the firing of anyone found to be gay or lesbian in the U.S. military. The figure cited this time, just over $190 million to discharge 3,364 service members (at a cost of about $53,000 each) from FY2004 – FY2009, has been erroneously reported by mainstream and online media as “the cost” of DADT.
While some media outlets disclaim that this figure purports to only cover FY2004 – FY2009, none is passing along the numerous other disclaimers that the GAO included in its report, especially its inability to include certain needed data in the analysis.
More concerning, however, is what is not disclaimed in the GAO report. The data source for the GAO researchers is the Defense Manpower Data Center (DMDC) in Seaside, CA. However, DMDC does not maintain comprehensive personnel records on all service members, only on those whose activities voluntarily report their personnel data to DMDC.
The GAO’s methodology in calculating the costs associated with DADT has been exposed as flawed in the past. The 2005 GAO report on this issue was meticulously picked apart and reworked by a Blue Ribbon Commission based at the University of California at Santa Barbara later that year. Furthermore, past GAO reports on gays in the military, dating back to the early 1990s, were known to contain surprisingly derogatory language, and even more recent versions continue to be peppered with the talking points of those opposed to repealing DADT.
Researchers and writers inevitably have their own biases that sometimes find their way into the works that bear their name. But the system’s integrity is maintained by full public disclosure of those biases. I, for example, serve as the Executive Director of the pro-repeal organization Servicemembers United, as my bio clearly indicates. The biases of the GAO’s researchers, however, have never been fully disclosed. Instead, we only have their flawed research products as evidence of a larger problem within the GAO.