Sorry Edgar but Trillian is right you need to look at it in the context of total numbers.
Say 100 women and 100 men all commited a shoplifting offence and got sent to prison.
Now lets say that there are another 100 men who commited violent offences and got sent to prison but that there were no women sent to prison for violent offences.
This would show that 100% of women in prison were there for non-violent offences but only 50% of men were in prison for non-violent offences but the women in this case are not being treated unfairly.
You also said that "also twice as many women (27%) are in jail after their first offence - than men."
but this suffers from the same interpretation error - all that says to me is that men are more likely to re-offend.
All it tells you is, of the total prison population, 27% of women of first offenders (compared with 12% of men). What it doesn't tell you is how many of the 88% of men that are in for subsequent offences were also sent down for their first offence (75% of the male prison population are in for 3 or more offence compared with only 56% of the female population suggesting that women are less likely to re-offend).
The current stats here don't appear to break down sentences by sex and previous convictions, although the PDF that JeaninePattibone links to above, has some old data which actually shows for shoplifting, drugs and violent offences, less women than men get sent to prison whether first offenders or not:
8% or Male first offenders sent to prison for shoplifting compared with 1% of women and 15% of male reoffenders compared with 5% female.