From a new paper by Siminski, Ville, and Paull (January 2013).
The authros conclude:
Our results have caveats and thus may not necessarily apply in other contexts. In particular, the criminal offense data only cover the study population’s middle-age years. While young males have higher crime rates, we believe that our estimates are precise enough to be meaningful. Nevertheless, we cannot test for crime effects that occur shortly after army service. However, we can rule out any substantial ‘permanent’ effects.
Of course these results are also contingent on the specific context of Australian conscript training. It took place for a period of two years at most, and among men who were mainly 20 years of age when they enlisted. Training involved character shaping with positive as well as potentially negative elements. Training skills were wide, varied, and often unrelated to acts of combat. Even among those not deployed to Vietnam, recruits were prepared for a war of covert counter-insurgency rather than for large scale battles. This training may have been less realistic or desensitizing than in the US, because live rounds and ‘Quick Kill’ training techniques were not used. Australian training also included ‘softer’ pacification skills, motivated by the need for close cooperation with Vietnamese civilians in Australia’s role in the war.