It also seems like a very high fraction! According to a quick google, 1.8m/330m ~= 0.5% of the US population is in prison or jail. Typically when people say ‘loved one[s]’ they mean close relatives; presumably that is generally fewer than 50 people. So even if they were non-overlapping (which seems unlikely as criminality/incarceration runs in families) I’d expect this to apply to fewer than 1⁄4 women.
I tried to track down the source of this stat. It appears it might (?) come from Hedwig (2015)‘s analysis of data from 2006, which in turn attributes it to ‘figure not shown’. I tried to reverse engineer the stat from table 3, which suggests that (6,363,170+11,226,655)/(14,461,749+93,555,459) = 16% of women have a family member in prison (though it’s possible I mis-read the table). I think there is something strange with the data; even though humans have basically gender-balanced families only 9% of men reported having a family member in prison! Additionally, more black women report having a family member in prison than ‘anyone they are acquainted with’, even though presumably they are acquainted with their family. It’s possibly people were using very generous definitions of who is a family member—e.g. having a second cousin you never talk to in prison—but then I am skeptical these could reasonably be described as ‘loved ones’.
It also seems like a very high fraction! According to a quick google, 1.8m/330m ~= 0.5% of the US population is in prison or jail. Typically when people say ‘loved one[s]’ they mean close relatives; presumably that is generally fewer than 50 people. So even if they were non-overlapping (which seems unlikely as criminality/incarceration runs in families) I’d expect this to apply to fewer than 1⁄4 women.
I tried to track down the source of this stat. It appears it might (?) come from Hedwig (2015)‘s analysis of data from 2006, which in turn attributes it to ‘figure not shown’. I tried to reverse engineer the stat from table 3, which suggests that (6,363,170+11,226,655)/(14,461,749+93,555,459) = 16% of women have a family member in prison (though it’s possible I mis-read the table). I think there is something strange with the data; even though humans have basically gender-balanced families only 9% of men reported having a family member in prison! Additionally, more black women report having a family member in prison than ‘anyone they are acquainted with’, even though presumably they are acquainted with their family. It’s possibly people were using very generous definitions of who is a family member—e.g. having a second cousin you never talk to in prison—but then I am skeptical these could reasonably be described as ‘loved ones’.