Okay, I get you now.
The answer is, you cannot tell from percentages alone. For either set actually.
Sentencing may be based on bias in the judiciary and certainly those figures are a reason and good indicator to look and see if bias is taking place.
But sentencing also depends upon circumstances of the crimes mitigating factors, the ability of the state to punish without undue burden to itself, and exactly what sort of sentences are being considered when drawing up those percentages, etc.
To determine whether or not it's discriminatory, you have to eliminate ALL other reasons for the discrepancy for sentencing OTHER THAN race or gender. You have to be able to say there is NO reason for this crime other than discrimination.
So ...be use specific examples. You have to eliminate the possibility that men serve longer sentences because they are stealing cars while women are more engaged with petty theft, that men more often use guns in the commission of theft while women use trickery, that the study did not exclude home confinement which may be applied to women more so that the state doesn't have to foster out their kids and pay for their care just to punish them, that the state is I'll prepared to dole out longer sentences because there are fewer women's prisoners and the state budget cannot afford to build more to address an unexpected rise in female criminality...that kind of thing.
A woman doctor in a town I lived in was convicted of murder or manslaughter. I forget which one. It wasn't first degree. Anyway, she got 5 years of house arrest and then 7 more years of probation.
Is that because men are discriminated against in judicial sentencing?
Or was it because she had a young daughter with Down's Syndrome who would have no one to care for her if her mother was sent to prison. Her father bailed when she "came out wrong like that". The man she killed was an assisted suicide of an elderly, sick, already terminal man. Some of the family was in on it and begged her to "help their father". She had no history of criminality. She wasn't likely to commit a crime again. A lot of people don't agree that euthanasia with consent is murder so there's a political and moral element here. It wasn't violent. There were no guns involved, and so forth and so on.
Compare that to an average guy who murders his wife by violently beating her to death and gets six years in prison but only serves two because men prisoners are more heavily paroled than women prisoners.
Who's really being more heavily punished?
One certainly has more comfortable confinement and is not isolated from her family. The other serves less time and the murder was very violent.
It's very easy to look at percentages like that and fall into equivocating or making snap judgements or erroneous statements.
That perfect tage doesn't determine discrimination. It's a starting point for investigating if it's there.
