As it happens, the relationship between desertion from service and the type of service is, in fact, opposite to the one stated by Gil - my dissertation database, including a sample of more than 850 cases of desertion for the years 2000-2002 confirms that the vast majority of deserters are not in combat service, but in other roles in the army. So having men in combat positions has nothing to do with gender percentage among deserters.

As to the gender of delinquents, it is true that the percentage of women deserters is minor, but this pattern holds true for female crime and delinquency in general. Why? well, that's a very good question. Ask three criminologists, you'll end up with nine different answers, at least.

(diving back to work on aforementioned database instead of spending time on internet.)