Much like Rolling Stone, Sen. Kirsten Gillibrand (D-NY) refuses to the let the facts — or lack thereof — get in the way of her preferred feminist narrative. In her case, it’s the claim that the military has it in for women. Instead of university administrators and fraternity members, she accuses Department of Defense officials of lies and cover-ups and implies that any female living near a base is at risk from sexual predators in uniform.

In the past, she’s used questionable survey results to generate support for legislation that would have stripped military commanders of their authority in sexual assault cases. Although the measure was universally opposed by DOD leaders and didn’t pass, she recently renewed her attack. Gillibrand was undeterred by the fact that the same category of responses she previously used to create the perception of a crisis showed an improvement of over 25% in the latest survey, or that victims have shown a greater willingness to report the incidents that do occur.

She deserves a little credit (very little…) for creativity. With the data working against her original line of “reasoning,” she was at least smart enough to chose a different — if even more flawed — metric to focus on this time around: sexual assaults on military spouses and civilian women who live near bases. She chooses a few isolated anecdotes to try and make the case that the military is incapable of treating women fairly. She also cites the Pentagon’s exclusion of these women in its surveys as evidence its hiding something. But as a Pentagon spokesman explained, “The department does not have standing authority to survey non-DOD civilian populations. However, federal surveys have found that the prevalence of sexual assault for non-DOD civilian women is statistically the same for military women and female spouses of military members.”

Translation: The Pentagon can’t do what Gillibrand wants it to, but it’s unnecessary anyway because there is already a pretty good idea what the numbers would say — and they don’t support Gillibrand’s argument.

Gillibrand bemoans that cases pursued in the military legal system are plagued by witnesses who decide to not cooperate, victims whose testimony is not considered credible, and inconsistent punishments. While those may (or may not) be valid observations, they aren’t unique to the military and could just as easily be used to describe similar cases tried in civilian courts.

One of her examples cites the case of a service member who allegedly assaulted a civilian female. The investigating officer identified too many inconsistencies in the victim’s story to pursue a sexual assault case, but ended up securing convictions on several lesser charges. Gillibrand puts officials like this in a no-win situation: She’ll bash them and question their integrity if they try to prosecute on the more serious charge and lose, and she’ll bash them and question their integrity if they get a conviction on charges that aren’t suitably severe for her.

Meanwhile, she ignores an interesting aspect of the report: No improvement was noted in the 20% of sexual assault incidents with male victims.

Sexual assault offenders should indeed be punished. That said, not everyone accused is actually guilty, though it appears the only way to make Gillibrand happy is for every male accused of any type of sexual assault to be convicted, regardless of what the evidence says.