Revelations - in a Herald investigation published last week - that 20 years of rape cases were written out of official data with the stroke of a pen are shocking.
The sex offences - thousands of them - were dismissed simply by labelling them "K3 - no crime". These were assaults put in the too-hard basket by police; ones in which there was limited evidence, in which the victims were intoxicated, or where consent was unclear - some 15 per cent of sex crimes.
The coding took place until as recently as 2013 - a decade after the issue was flagged by Victoria University criminologist Jan Jordan. It had pernicious effects; distorting statistics so it appeared fewer sexual assaults were being reported, keeping the "unresolved" rate artificially low, and (because the assaults were put in the same category as the small number of false rape complaints) making thousands of women look like liars.
It also let potentially thousands of perpetrators off the hook.
It is further evidence of what has been long known: that the true number of sex offences committed has been vastly different to the number reported, different again to the number investigated, to the number making it to court, and to the number resulting in a conviction.