# of Sex Crime Reports Don't = # Sex Crimes

The Associated Press has three different and wildly conflicting reports out in the space of a few hours about sexual assault rates in the US military. While all the news stories appear to be about the same Pentagon report, each frames the findings differently.

First, we are told, "Military Sees Fewer Sex Assault Reports." Apparently fewer men and women in uniform (6% and 30% respectively) had reported sexual assault in 2007 than in 2006, after increases over the past several years before that. But, the article goes on to point out, this is meaningless information since there is no way to know this is related to actual assault rates or the difficulties of reporting assaults. (A later article merely bore the headline "Military Women Report Harrassment" - also a bit misleading since men were also victims. A third released an hour later simply gives the numbers - a mere 2,688* (!) and avoids comparisons.)

What's the story here, AP? Is it that things are getting better, that they're still bad? (Fox News, for one, picked up the story and reframed it not about reported assaults, but actual assaults, concluding wrongly that the "Pentagon Report Says... Sexual Assaults Down in 2007."

My humble opinion: the real story is that the military should standardize the way it tabulates yearly reports. When rape reporting procedures are made more victim-friendly, rape reports typically increase for a period because victims don't feel th need to stay silent. However both reports and assaults are expected to decrease over time due to a deterrent environment.

But since good data is rarely gathered and policies not held constant over time, it is hard to know what works best. Efforts in the humanitarian arena, for example, to address gender-based violence in refugee camps suffer from this problem. The US military would be an excellent test case on such dynamics, if it would get its act together. Instead:

This is the fourth year the military has compiled detailed statistics on sexual assaults. The reporting methods have changed each year, complicating efforts to evaluate progress or to determine whether it is the actual assaults or the reporting that is going up or down.

Some other interesting items in the reports:

According to the documents, 1,516 reports involved the Army; 565 for the Air Force; 394 for the Navy; and 213 for the Marines.

Score one for the Marines. Puppy-killers they may all be (JK, Cleitus!), but looks like they are less likely than their compatriots to molest their own women.

Unwanted sexual advances among men were more common in the Navy, the 2006 gender relations survey reveals.

So much for banning women from serving on nuclear subs on account of 'close quarters.'

Some 1,040 completed probes resulted in no action, either due to insufficient evidence or because those responsible were civilians, foreigners or unidentified.

Hmm. Does this mean the military will prosecute its own but does not assist its personnel in prosecuting civilians?

*To put this number in perspective, in 1998 the rape rate per 100,000 in the US was 34.4 per 100,000 or .00034%; 2,688 as a proportion of our 2,885,200 troops is .00092%almost 3 times higher than the national average, which was at the time the highest among all countries that report such statistics. More rapes or better reporting? Currently, no way to know.